https://namforum.com/wiki/index.php/Special:RecentChanges Track the most recent changes to the wiki in this feed. en MediaWiki 1.46.0 Tue, 01 Sep 2026 09:59:36 GMT https://namforum.com/wiki/index.php?title=MediaWiki:Common.css&diff=36&oldid=31 https://namforum.com/wiki/index.php?title=MediaWiki:Common.css&diff=36&oldid=31 <p></p> <table style="background-color: #fff; color: #202122;" data-mw-interface=""> <col class="diff-marker" /> <col class="diff-content" /> <col class="diff-marker" /> <col class="diff-content" /> <tr class="diff-title" lang="en"> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">← Older revision</td> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">Revision as of 03:00, 1 September 2026</td> </tr><tr><td colspan="4" class="diff-multi" lang="en">(4 intermediate revisions by the same user not shown)</td></tr><tr><td colspan="2" class="diff-lineno" id="mw-diff-left-l1">Line 1:</td> <td colspan="2" class="diff-lineno">Line 1:</td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>/* CSS placed here will be applied to all skins */</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>/* CSS placed here will be applied to all skins */</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">/* NAMFORUM Legacy Vector - wider sidebar for logo */</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;"></ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">div#mw-panel {</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">    width: 240px;</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">}</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;"></ins></div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>#p-logo {</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>#p-logo {</div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>     width: 220px;</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>     width: 220px;</div></td></tr> <tr><td colspan="2" class="diff-lineno" id="mw-diff-left-l9">Line 9:</td> <td colspan="2" class="diff-lineno">Line 15:</td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>     background-repeat: no-repeat;</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>     background-repeat: no-repeat;</div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>     background-position: center;</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>     background-position: center;</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">}</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;"></ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">/* Move the Vector page structure right */</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">#mw-head-base,</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">div#content,</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">div#footer {</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">    margin-left: 240px;</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">}</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;"></ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">#left-navigation {</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">    left: 240px;</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">    margin-left: 1em;</ins></div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>}</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>}</div></td></tr> </table> Tue, 01 Sep 2026 07:00:32 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/MediaWiki_talk:Common.css https://namforum.com/wiki/index.php?title=MediaWiki:Common.css&diff=31&oldid=0 https://namforum.com/wiki/index.php?title=MediaWiki:Common.css&diff=31&oldid=0 <p>Created page with &quot;<span class="autocomment">CSS placed here will be applied to all skins: </span> #p-logo { width: 220px; } #p-logo a { width: 220px; background-size: 220px 75px; background-repeat: no-repeat; background-position: center; }&quot;</p> <p><b>New page</b></p><div>/* CSS placed here will be applied to all skins */<br /> #p-logo {<br /> width: 220px;<br /> }<br /> <br /> #p-logo a {<br /> width: 220px;<br /> background-size: 220px 75px;<br /> background-repeat: no-repeat;<br /> background-position: center;<br /> }</div> Tue, 01 Sep 2026 06:52:28 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/MediaWiki_talk:Common.css https://namforum.com/wiki/index.php?title=File:220-75.png&diff=30&oldid=0 https://namforum.com/wiki/index.php?title=File:220-75.png&diff=30&oldid=0 <p><a href="/wiki/index.php?title=User:NAMFORUM_Sysop&amp;action=edit&amp;redlink=1" class="mw-userlink new" title="User:NAMFORUM Sysop (page does not exist)"><bdi>NAMFORUM Sysop</bdi></a> uploaded <a href="/wiki/index.php/File:220-75.png" title="File:220-75.png">File:220-75.png</a></p> <p><b>New page</b></p><div></div> Tue, 01 Sep 2026 06:49:13 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/File_talk:220-75.png https://namforum.com/wiki/index.php?title=File:377-90.png&diff=29&oldid=0 https://namforum.com/wiki/index.php?title=File:377-90.png&diff=29&oldid=0 <p><a href="/wiki/index.php?title=User:NAMFORUM_Sysop&amp;action=edit&amp;redlink=1" class="mw-userlink new" title="User:NAMFORUM Sysop (page does not exist)"><bdi>NAMFORUM Sysop</bdi></a> uploaded <a href="/wiki/index.php/File:377-90.png" title="File:377-90.png">File:377-90.png</a></p> <p><b>New page</b></p><div></div> Tue, 01 Sep 2026 06:36:44 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/File_talk:377-90.png https://namforum.com/wiki/index.php?title=MediaWiki:Sidebar&diff=28&oldid=27 https://namforum.com/wiki/index.php?title=MediaWiki:Sidebar&diff=28&oldid=27 <p></p> <table style="background-color: #fff; color: #202122;" data-mw-interface=""> <col class="diff-marker" /> <col class="diff-content" /> <col class="diff-marker" /> <col class="diff-content" /> <tr class="diff-title" lang="en"> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">← Older revision</td> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">Revision as of 02:26, 1 September 2026</td> </tr><tr><td colspan="2" class="diff-lineno" id="mw-diff-left-l1">Line 1:</td> <td colspan="2" class="diff-lineno">Line 1:</td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>* navigation</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>* navigation</div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>** mainpage|mainpage-description</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>** mainpage|mainpage-description</div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>** https://namforum.com/|NAMFORUM <del style="font-weight: bold; text-decoration: none;">Forum</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div>** https://namforum.com/|<ins style="font-weight: bold; text-decoration: none;">Return to </ins>NAMFORUM</div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>** recentchanges-url|recentchanges</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>** recentchanges-url|recentchanges</div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>** randompage-url|randompage</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>** randompage-url|randompage</div></td></tr> </table> Tue, 01 Sep 2026 06:26:25 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/MediaWiki_talk:Sidebar https://namforum.com/wiki/index.php?title=MediaWiki:Sidebar&diff=27&oldid=0 https://namforum.com/wiki/index.php?title=MediaWiki:Sidebar&diff=27&oldid=0 <p>Modified sidebar navigation</p> <p><b>New page</b></p><div>* navigation<br /> ** mainpage|mainpage-description<br /> ** https://namforum.com/|NAMFORUM Forum<br /> ** recentchanges-url|recentchanges<br /> ** randompage-url|randompage<br /> ** helppage|help-mediawiki<br /> ** specialpages-url|specialpages<br /> <br /> * SEARCH<br /> <br /> * TOOLBOX<br /> <br /> * LANGUAGES</div> Tue, 01 Sep 2026 06:25:40 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/MediaWiki_talk:Sidebar https://namforum.com/wiki/index.php?title=Meta:History&diff=26&oldid=0 https://namforum.com/wiki/index.php?title=Meta:History&diff=26&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>== Origins ==<br /> <br /> NAMFORUM was founded by [[MJ Klein]] in 2026.<br /> <br /> The idea for the community grew out of Klein&#039;s own experiments with neural amp modeling and capture technology. After creating and publishing several models, he was surprised by how rapidly some of his tones were downloaded, shared, and used by other musicians.<br /> <br /> The response prompted a closer look at the rapidly developing field surrounding [[Neural Amp Modeler]] and related capture technologies. What initially appeared to be an interesting modeling technology was clearly becoming a much larger ecosystem of users, model creators, software developers, hardware manufacturers, and new approaches to capturing and reproducing audio equipment.<br /> <br /> Researching the field also revealed a need that went beyond simply providing another place to exchange models. As the technology grew, new users needed somewhere to ask questions, understand how capture technology works, learn proper techniques, troubleshoot problems, compare different approaches, and benefit from the experience of others.<br /> <br /> NAMFORUM was created to help support that growth, with &#039;&#039;&#039;education&#039;&#039;&#039; as one of its central purposes.<br /> <br /> The scope was deliberately made broader than Neural Amp Modeler itself. [[TONEX]], [[Kemper Profiling|Kemper PROFILER]], [[Neural Capture]], [[Line 6 Proxy]], and open-source projects such as AIDA-X and GuitarML all approach capture, profiling, or neural modeling in different ways. Bringing users of these technologies together provides opportunities to learn not only how a particular product works, but also the principles that apply across platforms.<br /> <br /> That educational purpose also led to the decision to build the &#039;&#039;&#039;NAMFORUM Wiki&#039;&#039;&#039; alongside the discussion forum. The forum provides a place to ask questions, experiment, exchange experience, and develop knowledge; the Wiki provides a place to organize and preserve what has been learned.</div> Tue, 01 Sep 2026 06:22:14 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Meta:History https://namforum.com/wiki/index.php?title=Meta:Administrators&diff=25&oldid=0 https://namforum.com/wiki/index.php?title=Meta:Administrators&diff=25&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>== About MJ Klein ==<br /> <br /> &#039;&#039;&#039;MJ Klein&#039;&#039;&#039; is the founder and administrator of NAMFORUM.<br /> <br /> He is a recording and mixing engineer with decades of experience in professional audio, recording technology, musical instruments, and electronic music equipment. His work includes recording, mixing, mastering, system integration, equipment testing, and the practical application of both hardware and software audio technology.<br /> <br /> Klein also spent approximately four decades working in manufacturing and engineering. That background strongly influences his approach to audio technology: understanding how a system works, establishing repeatable processes, testing results, documenting what was actually done, and distinguishing measurable behavior from assumptions or marketing claims.<br /> <br /> He has worked directly with manufacturers and software developers in testing equipment, evaluating software and firmware, documenting problems, and providing technical feedback. His current work includes hands-on experimentation with neural amp modeling and capture technologies, including the creation, testing, and comparison of models and capture workflows.<br /> <br /> Klein is also the administrator of &#039;&#039;&#039;Behringer World&#039;&#039;&#039;, an independent technical community and knowledge base devoted to Behringer digital audio equipment. His role there has included maintaining the forum and Wiki infrastructure and preserving community-developed technical knowledge, including overseeing the migration of the original DokuWiki knowledge base to MediaWiki.<br /> <br /> NAMFORUM grew from the same basic philosophy: useful technical knowledge should not disappear into social-media feeds, closed groups, or thousands of disconnected forum posts. Discussion belongs in the forum; information that proves useful over time can be organized and preserved in the Wiki.<br /> <br /> NAMFORUM is independent and is not operated by, sponsored by, or officially affiliated with Neural Amp Modeler or any manufacturer of capture, profiling, or modeling technology.</div> Tue, 01 Sep 2026 06:18:13 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Meta:Administrators https://namforum.com/wiki/index.php?title=Meta:Contributors&diff=24&oldid=0 https://namforum.com/wiki/index.php?title=Meta:Contributors&diff=24&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Contributors =<br /> <br /> The &#039;&#039;&#039;NAMFORUM Wiki&#039;&#039;&#039; is a community knowledge base for Neural Amp Modeling technology and other capture technology.<br /> <br /> == Current contributors ==<br /> <br /> The initial NAMFORUM Wiki was created and is maintained by &#039;&#039;&#039;MJ Klein&#039;&#039;&#039;, founder and administrator of NAMFORUM.<br /> <br /> Much of the initial Wiki content was developed with substantial research and drafting assistance from &#039;&#039;&#039;OpenAI&#039;s ChatGPT&#039;&#039;&#039;. Technical information is reviewed and, where practical, checked against primary sources including official documentation, developer publications, software repositories, manuals, and release notes before publication.<br /> <br /> AI-generated material is not considered authoritative merely because it was generated by AI. Responsibility for what is published in the Wiki remains with the Wiki&#039;s human editors and administrators.<br /> <br /> == Editorial approach ==<br /> <br /> The purpose of the Wiki is to document useful and reasonably established information about capture technology.<br /> <br /> Sources are generally preferred in the following order:<br /> <br /> * Official documentation and manuals<br /> * Original developer or manufacturer information<br /> * Source-code repositories<br /> * Original research and technical publications<br /> * Reproducible measurements and testing<br /> * Well-documented community experience<br /> <br /> Forum discussion, personal experience, and experimentation can identify information worth documenting, but claims should be distinguished from established technical facts.<br /> <br /> == Community contributions ==<br /> <br /> Public editing is not currently enabled.<br /> <br /> NAMFORUM members are nevertheless encouraged to contribute knowledge through discussion on the forum.<br /> <br /> Useful contributions may include:<br /> <br /> * Corrections<br /> * Technical information<br /> * New resources<br /> * Capture techniques<br /> * Measurements and testing<br /> * Documentation of new platforms or software<br /> * Suggestions for new Wiki articles<br /> * Identification of outdated information<br /> <br /> Information developed through community discussion may be incorporated into the Wiki by its editors.<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Visit NAMFORUM to participate in the community]&#039;&#039;&#039;<br /> <br /> == Attribution ==<br /> <br /> Where a contributor provides substantial original research, measurements, documentation, or other material, appropriate attribution may be included in the relevant article or its revision history.<br /> <br /> The MediaWiki revision history also provides a permanent record of edits made to each page.<br /> <br /> == See also ==<br /> <br /> * [[Meta:Administrators]]<br /> * [[Meta:History]]<br /> * [[Resources]]<br /> * [[Capture Technologies]]<br /> <br /> [[Category:NAMFORUM Wiki]]<br /> [[Category:Meta]]</div> Tue, 01 Sep 2026 06:16:04 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Meta:Contributors https://namforum.com/wiki/index.php?title=Resources&diff=23&oldid=0 https://namforum.com/wiki/index.php?title=Resources&diff=23&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Resources =<br /> <br /> This page collects useful resources for &#039;&#039;&#039;capture, profiling, neural audio modeling, model training, playback, and model sharing&#039;&#039;&#039;.<br /> <br /> Where possible, NAMFORUM Wiki favors:<br /> <br /> * Official documentation<br /> * Primary project sources<br /> * Original developer repositories<br /> * Official training tools<br /> * Established model libraries<br /> <br /> Third-party resources can be added when they provide substantial technical or practical value.<br /> <br /> Because capture technology develops rapidly, software versions, hardware support, model counts, and online services may change.<br /> <br /> __TOC__<br /> <br /> == Neural Amp Modeler ==<br /> <br /> === Official NAM website ===<br /> <br /> &#039;&#039;&#039;Neural Amp Modeler&#039;&#039;&#039;<br /> <br /> [https://www.neuralampmodeler.com/ neuralampmodeler.com]<br /> <br /> The primary website for the Neural Amp Modeler project.<br /> <br /> It provides access to:<br /> <br /> * NAM software<br /> * Training resources<br /> * Project information<br /> * Developer resources<br /> * Community links<br /> <br /> See [[Neural Amp Modeler]].<br /> <br /> === NAM user resources ===<br /> <br /> [https://www.neuralampmodeler.com/users NAM for Users]<br /> <br /> The official starting point for NAM users.<br /> <br /> It includes links for:<br /> <br /> * NAM model creation<br /> * Google Colab training<br /> * Local training<br /> * Standardized training files<br /> * Gateway<br /> * Model sharing<br /> <br /> === NAM documentation ===<br /> <br /> [https://neural-amp-modeler.readthedocs.io/ Neural Amp Modeler documentation]<br /> <br /> Technical documentation for the NAM training system and model format.<br /> <br /> Important sections include:<br /> <br /> * Training<br /> * Local GUI<br /> * Calibration<br /> * Model file specification<br /> * Metadata<br /> * Development information<br /> <br /> See:<br /> <br /> * [[Training a NAM Model]]<br /> * [[NAM Model File Format]]<br /> * [[Gain Staging and Calibration]]<br /> <br /> === NAM source code ===<br /> <br /> [https://www.neuralampmodeler.com/the-code NAM — The Code]<br /> <br /> The NAM project maintains separate open-source components for:<br /> <br /> * Trainer<br /> * Core DSP<br /> * Plugin<br /> <br /> These components make it possible to study and implement NAM training and playback independently.<br /> <br /> === Gateway ===<br /> <br /> &#039;&#039;&#039;Gateway&#039;&#039;&#039; is the official NAM model player linked from the NAM user resources.<br /> <br /> Use the current download links at:<br /> <br /> [https://www.neuralampmodeler.com/users NAM for Users]<br /> <br /> Gateway provides desktop playback of NAM snapshot models.<br /> <br /> See [[NAM Playback]].<br /> <br /> === TONE3000 ===<br /> <br /> [https://tone3000.com/ TONE3000]<br /> <br /> TONE3000 is a community-organized online model library and training service used extensively by the NAM community.<br /> <br /> The official NAM website directs users to TONE3000 for model sharing.<br /> <br /> TONE3000 provides:<br /> <br /> * NAM model discovery<br /> * Creator profiles<br /> * Model sharing<br /> * Online training<br /> * Search and organization<br /> <br /> Because online services evolve rapidly, consult TONE3000 directly for current supported model types and training features.<br /> <br /> == TONEX ==<br /> <br /> === Official TONEX website ===<br /> <br /> [https://www.ikmultimedia.com/products/tonex/ IK Multimedia TONEX]<br /> <br /> The primary source for the TONEX ecosystem.<br /> <br /> TONEX includes:<br /> <br /> * TONEX Player<br /> * TONEX Modeler<br /> * TONEX Editor<br /> * ToneNET<br /> * TONEX hardware<br /> * Mobile integration<br /> <br /> See [[TONEX]].<br /> <br /> === TONEX Modeler ===<br /> <br /> TONEX Modeler is IK Multimedia&#039;s dedicated application for creating Tone Models.<br /> <br /> Current information is available through:<br /> <br /> [https://www.ikmultimedia.com/products/tonex/ TONEX]<br /> <br /> TONEX Modeler supports modeling of equipment including:<br /> <br /> * Amplifiers<br /> * Cabinets<br /> * Combos<br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> * EQ<br /> * Boost<br /> * Complete compatible rigs<br /> <br /> === ToneNET ===<br /> <br /> [https://www.tone.net/tonex ToneNET]<br /> <br /> IK Multimedia&#039;s online service for discovering and sharing TONEX content.<br /> <br /> ToneNET provides access to community-created Tone Models and presets.<br /> <br /> Exact library counts change continuously and should not be treated as permanent specifications.<br /> <br /> === TONEX hardware ===<br /> <br /> Current TONEX hardware information is available through:<br /> <br /> [https://www.ikmultimedia.com/products/tonex/ TONEX]<br /> <br /> The ecosystem includes dedicated hardware for portable and live Tone Model playback.<br /> <br /> Consult current IK Multimedia documentation for compatibility and feature differences between devices.<br /> <br /> == Kemper Profiling ==<br /> <br /> === Official Kemper website ===<br /> <br /> [https://www.kemper-amps.com/ Kemper]<br /> <br /> The primary source for Kemper PROFILER hardware, software, documentation, and current profiling technology.<br /> <br /> See [[Kemper Profiling]].<br /> <br /> === PROFILER information ===<br /> <br /> [https://www.kemper-amps.com/profiler/overview Kemper PROFILER]<br /> <br /> Information about the current PROFILER ecosystem, including:<br /> <br /> * PROFILING<br /> * Liquid Profiling<br /> * PROFILING 2.0<br /> * PROFILER hardware<br /> * Effects<br /> * Rig management<br /> <br /> === Rig Manager ===<br /> <br /> &#039;&#039;&#039;Rig Manager&#039;&#039;&#039; is Kemper&#039;s software for managing PROFILER content.<br /> <br /> It provides access to:<br /> <br /> * Rigs<br /> * Presets<br /> * PROFILER management<br /> * Rig Exchange<br /> <br /> Current downloads and documentation are available through the Kemper website.<br /> <br /> === Rig Exchange ===<br /> <br /> &#039;&#039;&#039;Rig Exchange&#039;&#039;&#039; is Kemper&#039;s community library for shared Rigs.<br /> <br /> Access is available through Rig Manager and the Kemper ecosystem.<br /> <br /> Because the library is continuously changing, exact Rig counts should be treated as time-sensitive.<br /> <br /> === Kemper documentation ===<br /> <br /> Current manuals and support documentation are available through:<br /> <br /> [https://www.kemper-amps.com/downloads Kemper Downloads]<br /> <br /> For technical questions, current Kemper documentation should take precedence over older tutorials or forum posts.<br /> <br /> == Neural DSP Neural Capture ==<br /> <br /> === Quad Cortex ===<br /> <br /> [https://neuraldsp.com/quad-cortex Neural DSP Quad Cortex]<br /> <br /> Official information about Neural DSP&#039;s Quad Cortex platform.<br /> <br /> Quad Cortex combines:<br /> <br /> * Neural Capture<br /> * Algorithmic amplifier modeling<br /> * Cabinet processing<br /> * Effects<br /> * Routing<br /> * Hardware control<br /> <br /> See [[Neural Capture]].<br /> <br /> === Quad Cortex manual ===<br /> <br /> [https://neuraldsp.com/manual/quad-cortex Quad Cortex User Manual]<br /> <br /> The primary technical reference for:<br /> <br /> * Neural Capture<br /> * Cortex routing<br /> * Connections<br /> * Capture procedures<br /> * Presets<br /> * Device operation<br /> <br /> === Cortex Control ===<br /> <br /> [https://neuraldsp.com/cortex-control Cortex Control]<br /> <br /> Neural DSP&#039;s desktop application for Cortex hardware.<br /> <br /> Cortex Control also participates in the Neural Capture V2 creation workflow.<br /> <br /> === Cortex Cloud ===<br /> <br /> &#039;&#039;&#039;Cortex Cloud&#039;&#039;&#039; is Neural DSP&#039;s online system for Cortex content.<br /> <br /> It is used for:<br /> <br /> * Neural Capture sharing<br /> * Content discovery<br /> * User libraries<br /> * Neural Capture V2 cloud processing<br /> <br /> Access is provided through Neural DSP&#039;s Cortex ecosystem.<br /> <br /> === Neural Capture V2 ===<br /> <br /> Current information about Neural Capture V2 should be obtained directly from Neural DSP.<br /> <br /> See:<br /> <br /> [https://neuraldsp.com/news/introducing-neural-capture-version-2 Introducing Neural Capture Version 2]<br /> <br /> See also [[Neural Capture]].<br /> <br /> == Line 6 Proxy ==<br /> <br /> === Helix Stadium ===<br /> <br /> [https://line6.com/helix-stadium/ Helix Stadium]<br /> <br /> The primary Line 6 product resource for the Helix Stadium platform.<br /> <br /> Stadium combines:<br /> <br /> * Line 6 amplifier modeling<br /> * Agoura modeling<br /> * Proxy Clones<br /> * Cabinet processing<br /> * Effects<br /> * Advanced routing<br /> <br /> See [[Line 6 Proxy]].<br /> <br /> === Proxy Cloning documentation ===<br /> <br /> [https://manuals.line6.com/en/helix-stadium/live/clones Proxy Cloning]<br /> <br /> The current Line 6 documentation for creating and using Proxy Clones.<br /> <br /> It covers:<br /> <br /> * Amp+Cab Clones<br /> * Amp Clones<br /> * Preamp Clones<br /> * Distortion Clones<br /> * Connections<br /> * Level setup<br /> * Cloud processing<br /> * Clone management<br /> <br /> === Helix Stadium manual ===<br /> <br /> [https://manuals.line6.com/en/helix-stadium/live Helix Stadium Owner&#039;s Manual]<br /> <br /> The primary technical reference for Stadium hardware and software operation.<br /> <br /> === CustomTone ===<br /> <br /> [https://line6.com/customtone/ Line 6 CustomTone]<br /> <br /> Line 6&#039;s community service for sharing compatible presets and Proxy Clones.<br /> <br /> == Open capture technologies ==<br /> <br /> Open-source capture technology extends beyond NAM.<br /> <br /> See [[Open Capture Technologies]].<br /> <br /> === AIDA-X ===<br /> <br /> [https://github.com/AidaDSP/AIDA-X AIDA-X]<br /> <br /> AIDA-X is an open-source Amp Model player developed by Aida DSP.<br /> <br /> The project supports neural models of:<br /> <br /> * Amplifiers<br /> * Cabinets<br /> * Drive pedals<br /> * Fuzz<br /> * Boost<br /> * EQ<br /> * Combined signal chains<br /> <br /> AIDA-X uses RTNeural for neural-network inference and provides desktop and embedded implementations.<br /> <br /> === GuitarML Proteus ===<br /> <br /> [https://github.com/GuitarML/Proteus GuitarML Proteus]<br /> <br /> Proteus is an open-source guitar amplifier and pedal capture system.<br /> <br /> It supports:<br /> <br /> * Snapshot models<br /> * Conditioned models<br /> * Amplifier captures<br /> * Pedal captures<br /> * Plugin captures<br /> * Combined rigs<br /> <br /> The project includes a capture utility and training workflow.<br /> <br /> === GuitarML NeuralPi ===<br /> <br /> [https://github.com/GuitarML/NeuralPi NeuralPi]<br /> <br /> An open-source GuitarML project exploring neural guitar processing and embedded model playback.<br /> <br /> === GuitarML ToneLibrary ===<br /> <br /> [https://github.com/GuitarML/ToneLibrary GuitarML ToneLibrary]<br /> <br /> A library of models for GuitarML projects.<br /> <br /> Model compatibility depends on the neural architecture and intended player.<br /> <br /> Do not assume that every GuitarML JSON model is compatible with every GuitarML application.<br /> <br /> === SmartGuitarAmp ===<br /> <br /> [https://github.com/GuitarML/SmartGuitarAmp SmartGuitarAmp]<br /> <br /> An earlier GuitarML neural amplifier project using WaveNet-style modeling.<br /> <br /> It is useful both historically and technically for understanding the development of open neural guitar modeling.<br /> <br /> == Model libraries ==<br /> <br /> Capture technology becomes substantially more useful when models can be shared.<br /> <br /> Major sharing ecosystems include:<br /> <br /> {| class=&quot;wikitable&quot;<br /> ! Platform<br /> ! Model library / sharing service<br /> |-<br /> | [[Neural Amp Modeler]]<br /> | [https://tone3000.com/ TONE3000]<br /> |-<br /> | [[TONEX]]<br /> | [https://www.tone.net/tonex ToneNET]<br /> |-<br /> | [[Kemper Profiling]]<br /> | Rig Exchange<br /> |-<br /> | [[Neural Capture]]<br /> | Cortex Cloud<br /> |-<br /> | [[Line 6 Proxy]]<br /> | [https://line6.com/customtone/ CustomTone]<br /> |-<br /> | GuitarML<br /> | [https://github.com/GuitarML/ToneLibrary ToneLibrary]<br /> |}<br /> <br /> Availability, account requirements, and supported model types can change.<br /> <br /> == Training resources ==<br /> <br /> Before creating a model, see:<br /> <br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> For NAM specifically:<br /> <br /> * [[Training a NAM Model]]<br /> * [[NAM Model File Format]]<br /> * [[NAM A2 Architecture]]<br /> <br /> == Capture hardware ==<br /> <br /> Capture workflows may require additional equipment beyond the model-training software.<br /> <br /> Depending on the target, this can include:<br /> <br /> * Audio interface<br /> * Reamp box<br /> * DI box<br /> * Load box<br /> * Microphone<br /> * Microphone preamp<br /> * Appropriate cables<br /> * Measurement equipment<br /> <br /> No single hardware configuration is required for every capture.<br /> <br /> The correct signal chain depends on what is being modeled.<br /> <br /> See [[Capture Signal Chain]].<br /> <br /> == Reamping resources ==<br /> <br /> Reamping is fundamental to many computer-based capture workflows.<br /> <br /> See:<br /> <br /> * [[Reamping for Capture]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Capture Signal Chain]]<br /> <br /> The objective is to send a repeatable training signal through physical equipment while maintaining appropriate levels and connections.<br /> <br /> == Load boxes ==<br /> <br /> Direct amplifier capture generally requires an appropriate load when working from a power-amplifier speaker output.<br /> <br /> A load box may provide:<br /> <br /> * Safe amplifier load<br /> * Line-level output<br /> * Attenuation<br /> * Reactive loading<br /> * Cabinet simulation<br /> * IR processing<br /> <br /> When making a direct amplifier capture, cabinet or speaker simulation in the load box should normally be disabled unless intentionally included in the model.<br /> <br /> &#039;&#039;&#039;Never connect an amplifier speaker output directly to an ordinary audio-interface or capture-device input.&#039;&#039;&#039;<br /> <br /> == Cabinet impulse responses ==<br /> <br /> Cabinet IRs are widely used with direct amplifier captures.<br /> <br /> An IR can represent the approximately linear response of a:<br /> <br /> &#039;&#039;&#039;speaker + cabinet + microphone&#039;&#039;&#039;<br /> <br /> chain.<br /> <br /> Useful background is available at:<br /> <br /> [[Cabinets and IRs]]<br /> <br /> When downloading an IR, check:<br /> <br /> * Sample rate<br /> * File format<br /> * Length<br /> * Microphone<br /> * Speaker<br /> * Cabinet<br /> * Microphone position<br /> <br /> where that information is available.<br /> <br /> == Technical research ==<br /> <br /> Capture technology overlaps with research fields including:<br /> <br /> * Nonlinear system identification<br /> * Neural audio processing<br /> * Deep learning<br /> * Recurrent neural networks<br /> * Temporal convolutional networks<br /> * Real-time neural inference<br /> * Differentiable digital signal processing<br /> <br /> Academic papers can provide valuable information about the architectures and evaluation methods behind modern capture systems.<br /> <br /> However, research terminology and product terminology are not always equivalent.<br /> <br /> A commercial feature described as a &#039;&#039;capture&#039;&#039;, &#039;&#039;profile&#039;&#039;, or &#039;&#039;clone&#039;&#039; should not be assumed to use a particular published neural architecture unless the developer has documented that architecture.<br /> <br /> == GitHub ==<br /> <br /> [https://github.com/ GitHub]<br /> <br /> Many open-source audio-modeling projects publish their source code and technical documentation on GitHub.<br /> <br /> Useful repositories include:<br /> <br /> * [https://github.com/AidaDSP/AIDA-X AIDA-X]<br /> * [https://github.com/GuitarML/Proteus GuitarML Proteus]<br /> * [https://github.com/GuitarML/NeuralPi GuitarML NeuralPi]<br /> * [https://github.com/GuitarML/ToneLibrary GuitarML ToneLibrary]<br /> <br /> NAM&#039;s current source-code links are maintained at:<br /> <br /> [https://www.neuralampmodeler.com/the-code NAM — The Code]<br /> <br /> == Evaluating third-party information ==<br /> <br /> Capture technology changes quickly.<br /> <br /> Before relying on a tutorial, video, forum post, or social-media comment, check:<br /> <br /> * Publication date<br /> * Software version<br /> * Hardware version<br /> * Model architecture<br /> * Whether the information applies to the current release<br /> * Whether the author actually performed the described procedure<br /> * Whether a primary source confirms the claim<br /> <br /> Older information can remain useful, but it may describe software or workflows that have since changed substantially.<br /> <br /> == Primary sources ==<br /> <br /> For factual Wiki articles, NAMFORUM should prefer primary sources where practical.<br /> <br /> Examples include:<br /> <br /> * Developer documentation<br /> * Official manuals<br /> * Source-code repositories<br /> * Release notes<br /> * Technical specifications<br /> * Original research papers<br /> * Statements from project developers<br /> <br /> Community discussion remains valuable for:<br /> <br /> * Practical experience<br /> * Troubleshooting<br /> * Comparisons<br /> * Workarounds<br /> * Experiments<br /> * Independent testing<br /> <br /> The two forms of information serve different purposes.<br /> <br /> == Contributing resources ==<br /> <br /> Useful resources may be suggested through NAMFORUM.<br /> <br /> Good candidates include:<br /> <br /> * Official documentation<br /> * Technical papers<br /> * Open-source projects<br /> * Training tools<br /> * Capture utilities<br /> * Model libraries<br /> * Technical measurements<br /> * Well-documented tutorials<br /> * Long-term archival resources<br /> <br /> A resource should provide meaningful value rather than simply being a promotional link.<br /> <br /> == NAMFORUM ==<br /> <br /> &#039;&#039;&#039;NAMFORUM&#039;&#039;&#039; is an independent discussion community covering Neural Amp Modeling technology and other capture technology.<br /> <br /> [https://namforum.com/ Visit NAMFORUM]<br /> <br /> The forum is the appropriate place to:<br /> <br /> * Ask questions<br /> * Discuss equipment<br /> * Share experience<br /> * Compare capture systems<br /> * Discuss training methods<br /> * Troubleshoot models<br /> * Suggest Wiki improvements<br /> * Discuss new capture technologies<br /> <br /> The Wiki is intended to preserve established technical information that emerges from documentation, research, and community experience.<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Glossary]]<br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> * [[Neural Amp Modeler]]<br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Neural Capture]]<br /> * [[Line 6 Proxy]]<br /> * [[Open Capture Technologies]]<br /> <br /> [[Category:Resources]]<br /> [[Category:Capture technology]]<br /> [[Category:Reference]]</div> Tue, 01 Sep 2026 06:14:01 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Resources https://namforum.com/wiki/index.php?title=Glossary&diff=22&oldid=0 https://namforum.com/wiki/index.php?title=Glossary&diff=22&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Glossary =<br /> <br /> This glossary defines terminology commonly used in &#039;&#039;&#039;capture, profiling, neural audio modeling, reamping, and model playback&#039;&#039;&#039;.<br /> <br /> Terminology is not completely standardized across manufacturers and open-source projects. Different platforms may use words such as &#039;&#039;capture&#039;&#039;, &#039;&#039;profile&#039;&#039;, &#039;&#039;model&#039;&#039;, and &#039;&#039;clone&#039;&#039; for related but technically different processes.<br /> <br /> Where a term has a platform-specific meaning, that distinction is noted.<br /> <br /> __TOC__<br /> <br /> == A ==<br /> <br /> === A1 ===<br /> <br /> The original family of standard [[Neural Amp Modeler]] neural-network architectures, retrospectively named &#039;&#039;&#039;A1&#039;&#039;&#039; following the introduction of [[NAM A2 Architecture|A2]].<br /> <br /> A1 includes architectures previously known as Standard, Lite, Feather, and Nano.<br /> <br /> A1 models remain supported by current NAM software.<br /> <br /> === A2 ===<br /> <br /> The second-generation standard architecture for [[Neural Amp Modeler]], officially released in June 2026.<br /> <br /> A2 was developed to improve the relationship between modeling accuracy, computational requirements, training efficiency, and hardware implementation.<br /> <br /> See [[NAM A2 Architecture]].<br /> <br /> === A/B comparison ===<br /> <br /> Direct comparison between two signal paths.<br /> <br /> During capture work, this commonly means comparing:<br /> <br /> * Physical device versus trained model<br /> * Original recording versus model output<br /> * Two different models of the same equipment<br /> <br /> Accurate A/B comparison requires careful level matching.<br /> <br /> === AIDA-X ===<br /> <br /> An open-source neural amplifier modeling and playback project developed by Aida DSP.<br /> <br /> AIDA-X uses its own model architecture and should not be confused with NAM merely because both are open-source neural modeling technologies.<br /> <br /> See [[Open Capture Technologies]].<br /> <br /> === Amp capture ===<br /> <br /> A model representing an amplifier without its physical speaker/cabinet/microphone response.<br /> <br /> An amp capture is commonly combined with a cabinet IR or cabinet model during playback.<br /> <br /> Also called a &#039;&#039;&#039;direct amp capture&#039;&#039;&#039; in some contexts.<br /> <br /> === Amp+Cab capture ===<br /> <br /> A capture that includes the amplifier together with a speaker/cabinet signal path.<br /> <br /> When created using a microphone, the microphone and its placement are also normally part of the resulting sound.<br /> <br /> An additional cabinet IR is generally unnecessary during conventional playback.<br /> <br /> See [[Capture Types]].<br /> <br /> === Algorithmic modeling ===<br /> <br /> Modeling in which a developer designs an algorithm intended to reproduce the behavior of a device or circuit.<br /> <br /> This contrasts conceptually with capture technology, where the behavior of physical equipment is measured and a model is derived from those measurements.<br /> <br /> Modern products may combine algorithmic and capture-based techniques.<br /> <br /> === Aliasing ===<br /> <br /> Unwanted frequencies produced when nonlinear processing generates frequency components above the Nyquist frequency and they fold back into the audible range.<br /> <br /> Oversampling is commonly used in nonlinear digital audio processing to reduce aliasing.<br /> <br /> === Architecture ===<br /> <br /> The structure of a neural network.<br /> <br /> An architecture defines how neural-network layers and processing elements are arranged.<br /> <br /> Examples include:<br /> <br /> * LSTM<br /> * WaveNet-style architectures<br /> * NAM A1<br /> * NAM A2<br /> <br /> Two model files using different architectures are not automatically interchangeable.<br /> <br /> === Audio interface ===<br /> <br /> Hardware that converts between analog audio signals and digital audio used by a computer.<br /> <br /> In a capture workflow, an audio interface may provide:<br /> <br /> * Training-signal output<br /> * Capture return input<br /> * Level control<br /> * Analog-to-digital conversion<br /> * Digital-to-analog conversion<br /> <br /> == B ==<br /> <br /> === Batch training ===<br /> <br /> Training multiple models as part of one automated workflow.<br /> <br /> Batch training can be useful when several recordings have been made from:<br /> <br /> * Different amplifier settings<br /> * Different pedals<br /> * Different channels<br /> * Different pieces of equipment<br /> <br /> === Bias ===<br /> <br /> An operating condition within an electronic circuit.<br /> <br /> In capture discussions, the term is often encountered with tube amplifiers or fuzz circuits whose behavior can change significantly according to bias conditions.<br /> <br /> Bias can also have unrelated meanings in machine learning. Context matters.<br /> <br /> === Block ===<br /> <br /> A processing element placed within a digital signal chain.<br /> <br /> Examples include:<br /> <br /> * Amp block<br /> * Capture block<br /> * Cab block<br /> * IR block<br /> * Delay block<br /> <br /> Several hardware capture platforms represent captured models as blocks within a larger preset.<br /> <br /> === Boost ===<br /> <br /> A device that increases signal level and may also alter frequency response or nonlinear behavior.<br /> <br /> Boost pedals are common capture targets.<br /> <br /> === Buffer ===<br /> <br /> An electronic circuit designed to isolate signal stages and provide appropriate impedance relationships.<br /> <br /> A buffer can alter the behavior of impedance-sensitive equipment such as certain vintage fuzz circuits.<br /> <br /> Its presence or absence can therefore matter during capture.<br /> <br /> == C ==<br /> <br /> === Cab ===<br /> <br /> Abbreviation for &#039;&#039;&#039;speaker cabinet&#039;&#039;&#039;.<br /> <br /> In modeling systems, &#039;&#039;Cab&#039;&#039; can also refer to a digital cabinet model.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> === Calibration ===<br /> <br /> The process of establishing a known relationship between digital signal level and analog signal level.<br /> <br /> In capture systems, calibration can help ensure that a model receives approximately the same operating level during playback that the physical equipment received during measurement.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> === Capture ===<br /> <br /> The measurement process used to characterize physical audio equipment for creation of a digital model.<br /> <br /> The word may also be used informally for the resulting model.<br /> <br /> Platform terminology varies.<br /> <br /> For example:<br /> <br /> * Neural DSP — Neural Capture<br /> * Line 6 — Clone<br /> * Kemper — PROFILE<br /> * TONEX — Tone Model<br /> * NAM — model<br /> <br /> === Capture data ===<br /> <br /> Recorded measurement information used to train or construct a model.<br /> <br /> Capture data is not necessarily the finished model.<br /> <br /> Conceptually:<br /> <br /> &#039;&#039;&#039;capture data → training/modeling process → model&#039;&#039;&#039;<br /> <br /> === Capture signal ===<br /> <br /> A specially prepared audio signal sent through physical equipment during the measurement process.<br /> <br /> Also called:<br /> <br /> * Training signal<br /> * Stimulus<br /> * Excitation signal<br /> * Test signal<br /> <br /> Different capture systems use different signals.<br /> <br /> === Capture signal chain ===<br /> <br /> The complete physical path through which the capture signal travels.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;interface → reamp box → amplifier → load box → interface&#039;&#039;&#039;<br /> <br /> or:<br /> <br /> &#039;&#039;&#039;capture hardware → amplifier → cabinet → microphone → capture hardware&#039;&#039;&#039;<br /> <br /> Everything within the measured signal path can potentially influence the resulting capture.<br /> <br /> See [[Capture Signal Chain]].<br /> <br /> === Capture type ===<br /> <br /> The portion of a signal chain represented by a model.<br /> <br /> Common capture types include:<br /> <br /> * Pedal<br /> * Amp<br /> * Amp+Cab<br /> * Preamp<br /> * Complete rig<br /> <br /> See [[Capture Types]].<br /> <br /> === Clone ===<br /> <br /> Line 6&#039;s term for a model created using [[Line 6 Proxy]].<br /> <br /> Current Proxy Clone categories include:<br /> <br /> * Amp+Cab<br /> * Amp<br /> * Preamp<br /> * Distortion<br /> <br /> The word &#039;&#039;clone&#039;&#039; may also be used generically in discussion, but &#039;&#039;&#039;Clone&#039;&#039;&#039; has a specific meaning within the Proxy ecosystem.<br /> <br /> === Cloud training ===<br /> <br /> Model training performed on remote computing infrastructure rather than entirely on the user&#039;s local computer or capture hardware.<br /> <br /> Examples include workflows used by:<br /> <br /> * Neural Capture V2<br /> * Line 6 Proxy<br /> <br /> Cloud training can provide greater computational resources but requires access to the associated online service.<br /> <br /> === Colab ===<br /> <br /> Common abbreviation for &#039;&#039;&#039;Google Colaboratory&#039;&#039;&#039;.<br /> <br /> Colab provides cloud-hosted Python notebook execution and has been widely used for neural-model training.<br /> <br /> NAM and several open-source audio-modeling projects have provided Colab-based training workflows.<br /> <br /> === Conditioned model ===<br /> <br /> A neural model that receives one or more control parameters in addition to the audio signal.<br /> <br /> Conceptually:<br /> <br /> &#039;&#039;&#039;audio + control value → model → output&#039;&#039;&#039;<br /> <br /> This can allow one model to represent behavior across multiple physical control positions.<br /> <br /> Contrast with [[#Snapshot model|snapshot model]].<br /> <br /> === Convolution ===<br /> <br /> A mathematical operation widely used in digital signal processing.<br /> <br /> Cabinet impulse-response playback typically uses convolution to apply the measured response contained in an IR to an audio signal.<br /> <br /> === Cortex Cloud ===<br /> <br /> Neural DSP&#039;s online service for Cortex content.<br /> <br /> It is used for sharing and managing Cortex material and performs the cloud-based training required for Neural Capture V2.<br /> <br /> === Cortex Control ===<br /> <br /> Neural DSP&#039;s desktop application for managing Cortex hardware.<br /> <br /> It also participates in the Neural Capture V2 creation workflow.<br /> <br /> === CPU ===<br /> <br /> &#039;&#039;&#039;Central Processing Unit.&#039;&#039;&#039;<br /> <br /> In capture playback, CPU requirements influence:<br /> <br /> * Number of models that can run simultaneously<br /> * Buffer size<br /> * Latency<br /> * Oversampling<br /> * Available effects processing<br /> <br /> Different neural architectures can have substantially different CPU requirements.<br /> <br /> === CustomTone ===<br /> <br /> Line 6&#039;s community service for sharing compatible presets and Proxy Clones.<br /> <br /> == D ==<br /> <br /> === dB ===<br /> <br /> &#039;&#039;&#039;Decibel.&#039;&#039;&#039;<br /> <br /> A logarithmic unit used to describe ratios such as signal level or gain.<br /> <br /> The meaning depends on the reference being used.<br /> <br /> === dBFS ===<br /> <br /> &#039;&#039;&#039;Decibels relative to Full Scale.&#039;&#039;&#039;<br /> <br /> A digital level measurement.<br /> <br /> 0 dBFS represents the maximum digital level available before numerical clipping in a conventional fixed reference system.<br /> <br /> Values below full scale are expressed as negative numbers.<br /> <br /> === dBu ===<br /> <br /> A voltage-based audio level measurement referenced to 0.775 volts RMS.<br /> <br /> Calibration information in some capture systems may use dBu to describe the relationship between digital and analog levels.<br /> <br /> === Device under test ===<br /> <br /> The physical equipment being measured.<br /> <br /> Often abbreviated &#039;&#039;&#039;DUT&#039;&#039;&#039;.<br /> <br /> Examples include:<br /> <br /> * Amplifier<br /> * Pedal<br /> * Preamp<br /> * Compressor<br /> * Complete rig<br /> <br /> === Direct capture ===<br /> <br /> A capture made without a physical speaker/microphone response.<br /> <br /> For an amplifier this commonly means taking the output through an appropriate load box or direct-output system.<br /> <br /> A cabinet model or IR is normally required when playing a direct amp capture through a full-range monitoring system.<br /> <br /> === Direct Injection box ===<br /> <br /> See [[#DI box|DI box]].<br /> <br /> === DI box ===<br /> <br /> &#039;&#039;&#039;Direct Injection box.&#039;&#039;&#039;<br /> <br /> A device commonly used to interface an instrument-level, high-impedance source with a balanced microphone-level input.<br /> <br /> A DI box and reamp box perform broadly opposite interface functions, although specific equipment may provide additional capabilities.<br /> <br /> === DSP ===<br /> <br /> &#039;&#039;&#039;Digital Signal Processing&#039;&#039;&#039; or, depending on context, the processing resources available for digital signal processing.<br /> <br /> Hardware modelers often describe limits in terms of available DSP.<br /> <br /> === DUT ===<br /> <br /> See [[#Device under test|Device under test]].<br /> <br /> == E ==<br /> <br /> === Embedded system ===<br /> <br /> A dedicated computing platform designed to perform a specific function.<br /> <br /> Neural-model players can run on embedded processors inside:<br /> <br /> * Pedals<br /> * Amplifiers<br /> * Multi-effects processors<br /> * Dedicated model players<br /> <br /> Efficient neural architectures are particularly important for embedded use.<br /> <br /> === Excitation signal ===<br /> <br /> See [[#Capture signal|Capture signal]].<br /> <br /> == F ==<br /> <br /> === Full-range system ===<br /> <br /> A playback system intended to reproduce a broad frequency range relatively neutrally.<br /> <br /> Capture users often use full-range monitors or PA speakers when playing models that already include cabinet simulation.<br /> <br /> === FRFR ===<br /> <br /> &#039;&#039;&#039;Full Range, Flat Response.&#039;&#039;&#039;<br /> <br /> A term commonly used for speaker systems intended to reproduce modeled guitar rigs without adding the strong tonal filtering of a traditional guitar cabinet.<br /> <br /> In practice, no loudspeaker is perfectly flat.<br /> <br /> === Fuzz ===<br /> <br /> A strongly nonlinear distortion effect.<br /> <br /> Some fuzz circuits are particularly challenging capture targets because their behavior can depend on:<br /> <br /> * Source impedance<br /> * Input level<br /> * Guitar volume<br /> * Bias<br /> * Playing dynamics<br /> <br /> == G ==<br /> <br /> === Gain staging ===<br /> <br /> Management of signal level throughout an audio chain.<br /> <br /> In capture work, gain staging is important both for avoiding clipping/noise and for reproducing the operating conditions under which nonlinear equipment was measured.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> === Ground lift ===<br /> <br /> A facility designed to interrupt an audio ground connection to help eliminate ground-loop hum.<br /> <br /> A proper audio ground lift should not be confused with defeating protective mains safety grounding.<br /> <br /> &#039;&#039;&#039;Never remove protective mains grounding as a method of eliminating hum.&#039;&#039;&#039;<br /> <br /> === Ground loop ===<br /> <br /> An unwanted current path created by multiple grounding connections.<br /> <br /> Ground loops commonly produce hum or other interference in capture and reamping systems.<br /> <br /> == H ==<br /> <br /> === Hardware capture ===<br /> <br /> A capture workflow in which dedicated hardware performs some or all of the measurement process.<br /> <br /> Examples include:<br /> <br /> * Kemper PROFILER<br /> * Quad Cortex<br /> * Helix Stadium<br /> <br /> The model training itself may occur locally or in the cloud depending on the platform.<br /> <br /> === Hidden layer ===<br /> <br /> An internal processing layer within a neural network.<br /> <br /> The size and organization of hidden layers are part of the model architecture.<br /> <br /> == I ==<br /> <br /> === Impedance ===<br /> <br /> Opposition to alternating current, including audio signals.<br /> <br /> Impedance relationships can affect the behavior of:<br /> <br /> * Guitar pickups<br /> * Pedals<br /> * Reamp devices<br /> * Amplifier inputs<br /> * Audio interfaces<br /> <br /> Some equipment, particularly certain fuzz circuits, can respond strongly to source impedance.<br /> <br /> === Impulse response ===<br /> <br /> See [[#IR|IR]].<br /> <br /> === Inference ===<br /> <br /> Using an already-trained neural network to process new input.<br /> <br /> In capture terminology:<br /> <br /> &#039;&#039;&#039;training&#039;&#039;&#039; creates the model.<br /> <br /> &#039;&#039;&#039;inference&#039;&#039;&#039; runs the model.<br /> <br /> Real-time capture playback is neural-network inference.<br /> <br /> === Input level ===<br /> <br /> The signal level presented to a device or model.<br /> <br /> With nonlinear equipment, input level affects behavior as well as output volume.<br /> <br /> Changing input level can alter:<br /> <br /> * Distortion<br /> * Compression<br /> * Sustain<br /> * Dynamics<br /> * Cleanup<br /> <br /> === IR ===<br /> <br /> &#039;&#039;&#039;Impulse Response.&#039;&#039;&#039;<br /> <br /> A representation of the response of a linear system.<br /> <br /> In guitar modeling, IRs are commonly used to represent:<br /> <br /> &#039;&#039;&#039;speaker + cabinet + microphone&#039;&#039;&#039;<br /> <br /> An IR is fundamentally different from a nonlinear neural amplifier model.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == J ==<br /> <br /> === JSON ===<br /> <br /> &#039;&#039;&#039;JavaScript Object Notation.&#039;&#039;&#039;<br /> <br /> A human-readable structured data format.<br /> <br /> Some neural model formats use JSON or JSON-based structures.<br /> <br /> The fact that two model formats use JSON does not make them compatible.<br /> <br /> The [[NAM Model File Format]] is JSON-based.<br /> <br /> == L ==<br /> <br /> === Latency ===<br /> <br /> Delay between an input signal entering a system and the corresponding output being produced.<br /> <br /> Latency matters in two different capture contexts:<br /> <br /> &#039;&#039;&#039;Recording/training&#039;&#039;&#039;<br /> <br /> The recorded output must be correctly aligned with the training input.<br /> <br /> &#039;&#039;&#039;Playback&#039;&#039;&#039;<br /> <br /> Excessive processing latency can make real-time playing uncomfortable.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> === Level matching ===<br /> <br /> Adjusting two signals so they have closely comparable perceived level.<br /> <br /> Level matching is essential when comparing:<br /> <br /> * Model versus physical device<br /> * Two models<br /> * Two capture platforms<br /> <br /> A slightly louder signal can easily be perceived as better.<br /> <br /> === Linear system ===<br /> <br /> A system whose behavior can be represented using linear signal-processing methods.<br /> <br /> Speaker/cabinet/microphone responses are often approximated as linear systems and represented using IRs.<br /> <br /> Distorting amplifiers and pedals are nonlinear and generally require more complex modeling.<br /> <br /> === Load box ===<br /> <br /> A device designed to provide the electrical load required by an amplifier while allowing its output to be recorded or processed without relying solely on a physical loudspeaker.<br /> <br /> Load boxes are commonly used when creating direct amplifier captures.<br /> <br /> Some load boxes also provide:<br /> <br /> * Speaker simulation<br /> * IR processing<br /> * Attenuation<br /> * Line outputs<br /> <br /> These functions should be configured appropriately for the intended capture type.<br /> <br /> === LSTM ===<br /> <br /> &#039;&#039;&#039;Long Short-Term Memory.&#039;&#039;&#039;<br /> <br /> A recurrent neural-network architecture designed to process sequential information.<br /> <br /> LSTM architectures have been used in several neural audio-modeling projects, including GuitarML systems.<br /> <br /> == M ==<br /> <br /> === Machine learning ===<br /> <br /> A broad class of computational methods in which a system learns relationships from data rather than relying solely on explicitly programmed rules.<br /> <br /> Several modern capture technologies use machine learning to reproduce nonlinear audio equipment.<br /> <br /> === Metadata ===<br /> <br /> Information stored with or associated with a model.<br /> <br /> Useful capture metadata may include:<br /> <br /> * Creator<br /> * Equipment manufacturer<br /> * Equipment model<br /> * Capture type<br /> * Equipment settings<br /> * Input calibration<br /> * Output calibration<br /> * Training information<br /> * Capture date<br /> <br /> See [[NAM Model File Format]] for NAM-specific metadata.<br /> <br /> === Mic capture ===<br /> <br /> A capture whose measured path includes a microphone.<br /> <br /> For an amplifier rig this usually means:<br /> <br /> &#039;&#039;&#039;amplifier → speaker → microphone&#039;&#039;&#039;<br /> <br /> The microphone and its position become part of the resulting model or captured response.<br /> <br /> === Model ===<br /> <br /> A digital representation used to reproduce the behavior of another system.<br /> <br /> Within capture technology, a model is the result of training or profiling measurements of physical equipment.<br /> <br /> The word is also used more broadly for conventional algorithmic amp and effects models.<br /> <br /> Context is important.<br /> <br /> === Model architecture ===<br /> <br /> See [[#Architecture|Architecture]].<br /> <br /> === Model conversion ===<br /> <br /> Transforming a model intended for one architecture or playback engine into another representation.<br /> <br /> Conversion is not necessarily lossless.<br /> <br /> A converted model should not automatically be described as executing the original architecture natively.<br /> <br /> See [[NAM Playback]].<br /> <br /> === Model file ===<br /> <br /> A file containing the information required by a compatible playback engine to run a trained model.<br /> <br /> A model file is not necessarily the same thing as the original capture recording.<br /> <br /> === Model player ===<br /> <br /> Software or hardware capable of executing a trained model.<br /> <br /> Examples range from desktop plugins to dedicated pedals and multi-effects processors.<br /> <br /> === Modeling ===<br /> <br /> The broad process of creating a computational representation of physical equipment.<br /> <br /> Modeling can include:<br /> <br /> * Circuit modeling<br /> * Behavioral modeling<br /> * Machine-learning modeling<br /> * Profiling<br /> * Capture technology<br /> <br /> == N ==<br /> <br /> === NAM ===<br /> <br /> &#039;&#039;&#039;Neural Amp Modeler.&#039;&#039;&#039;<br /> <br /> An open-source deep-learning system for modeling nonlinear audio equipment.<br /> <br /> NAM also commonly refers to the wider ecosystem surrounding the technology.<br /> <br /> See [[Neural Amp Modeler]].<br /> <br /> === .nam ===<br /> <br /> The conventional file extension for a [[Neural Amp Modeler]] model.<br /> <br /> A &#039;&#039;.nam&#039;&#039; file uses the [[NAM Model File Format]].<br /> <br /> Not every neural amplifier model is a NAM model.<br /> <br /> === NAM A1 ===<br /> <br /> See [[#A1|A1]].<br /> <br /> === NAM A2 ===<br /> <br /> See [[#A2|A2]].<br /> <br /> === Native playback ===<br /> <br /> Playback in which software or hardware executes the model architecture directly rather than first converting it to another proprietary model representation.<br /> <br /> See [[NAM Playback]].<br /> <br /> === Neural Capture ===<br /> <br /> Neural DSP&#039;s proprietary capture technology used in the Cortex ecosystem.<br /> <br /> Current generations include:<br /> <br /> * Neural Capture V1<br /> * Neural Capture V2<br /> <br /> See [[Neural Capture]].<br /> <br /> === Neural network ===<br /> <br /> A computational structure consisting of interconnected processing elements whose parameters are learned from data.<br /> <br /> Neural networks can approximate nonlinear relationships and are therefore useful for modeling audio equipment such as amplifiers and pedals.<br /> <br /> === Nonlinear system ===<br /> <br /> A system whose output is not simply a scaled or filtered version of its input.<br /> <br /> Examples include:<br /> <br /> * Distorting amplifier<br /> * Overdrive<br /> * Fuzz<br /> * Compressor<br /> * Saturating preamp<br /> <br /> Neural capture technology is particularly useful for reproducing nonlinear behavior.<br /> <br /> === Normalization ===<br /> <br /> Adjustment of signal or data level according to a defined reference.<br /> <br /> Normalization can mean different things in recording, training, and model playback contexts.<br /> <br /> It should not automatically be assumed to be equivalent to calibration.<br /> <br /> === Nyquist frequency ===<br /> <br /> Half the digital audio sample rate.<br /> <br /> For example, at a 48 kHz sample rate, the Nyquist frequency is 24 kHz.<br /> <br /> Frequency components above Nyquist cannot be represented directly at that sample rate.<br /> <br /> == O ==<br /> <br /> === Open source ===<br /> <br /> Software whose source code is available under a license permitting specified forms of use, inspection, modification, and redistribution.<br /> <br /> Open source does not mean:<br /> <br /> * No copyright<br /> * No license restrictions<br /> * Universal model compatibility<br /> * Free use for every possible purpose<br /> <br /> Always check the project&#039;s actual license.<br /> <br /> === Oversampling ===<br /> <br /> Processing audio internally at a higher sample rate than the system&#039;s nominal rate.<br /> <br /> Oversampling is commonly used in nonlinear audio processing to reduce audible aliasing.<br /> <br /> It increases computational requirements.<br /> <br /> == P ==<br /> <br /> === Pedal capture ===<br /> <br /> A model of a physical effects pedal.<br /> <br /> Capture technologies commonly support nonlinear pedals such as:<br /> <br /> * Boost<br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> <br /> Support for compressors and other dynamic processors varies by capture architecture.<br /> <br /> === Playback ===<br /> <br /> Running a completed model with new audio.<br /> <br /> For neural models this is inference.<br /> <br /> Playback requirements can differ substantially from training requirements.<br /> <br /> See [[NAM Playback]].<br /> <br /> === Preamp capture ===<br /> <br /> A model representing a preamplifier or the preamp section of an amplifier.<br /> <br /> A preamp capture normally excludes:<br /> <br /> * Power amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> <br /> unless those elements are deliberately included elsewhere in the measured path.<br /> <br /> === Profile ===<br /> <br /> Generic term sometimes used for a captured representation of equipment.<br /> <br /> &#039;&#039;&#039;PROFILE&#039;&#039;&#039; is also Kemper&#039;s specific term for the result of the Kemper Profiling process.<br /> <br /> See [[Kemper Profiling]].<br /> <br /> === Profiling ===<br /> <br /> A term for measuring equipment and creating a digital representation of its behavior.<br /> <br /> The term is particularly associated with Kemper, although it is also used generically.<br /> <br /> === Proteus ===<br /> <br /> An open-source neural-modeling project developed by GuitarML.<br /> <br /> Proteus supports snapshot and conditioned models.<br /> <br /> See [[Open Capture Technologies]].<br /> <br /> === Proxy ===<br /> <br /> Line 6&#039;s capture technology introduced for Helix Stadium.<br /> <br /> The resulting models are called &#039;&#039;&#039;Clones&#039;&#039;&#039;.<br /> <br /> See [[Line 6 Proxy]].<br /> <br /> == Q ==<br /> <br /> === Quad Cortex ===<br /> <br /> Neural DSP&#039;s Cortex hardware processor providing amplifier/effects modeling and Neural Capture functionality.<br /> <br /> It supports creation and playback of Neural Captures.<br /> <br /> == R ==<br /> <br /> === Reactive load ===<br /> <br /> A load device designed to present an amplifier with an electrical load whose impedance varies with frequency in a manner intended to resemble a loudspeaker system.<br /> <br /> Reactive load boxes are widely used for direct amplifier recording and capture.<br /> <br /> === Reamp box ===<br /> <br /> A device used to interface a line-level playback source with equipment designed for an instrument-level input.<br /> <br /> Typical applications include sending a recorded signal from an audio interface into:<br /> <br /> * Guitar amplifier<br /> * Pedal<br /> * Other instrument-level equipment<br /> <br /> A reamp box may provide:<br /> <br /> * Level attenuation<br /> * Balanced-to-unbalanced conversion<br /> * Ground isolation<br /> * Appropriate source impedance<br /> <br /> See [[Reamping for Capture]].<br /> <br /> === Reamping ===<br /> <br /> Sending previously recorded audio back through physical audio equipment.<br /> <br /> In capture workflows, reamping allows a standardized training signal to be sent repeatedly through different equipment or settings.<br /> <br /> See [[Reamping for Capture]].<br /> <br /> === Reference level ===<br /> <br /> A defined signal level used as a known measurement point.<br /> <br /> Reference levels are important when relating digital levels to analog voltage.<br /> <br /> === Return ===<br /> <br /> The signal coming back from the device under test into the capture or recording system.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;interface output → amplifier → load box → interface return&#039;&#039;&#039;<br /> <br /> === Rig ===<br /> <br /> A collection of equipment forming a usable signal chain.<br /> <br /> Depending on context, a rig may include:<br /> <br /> * Pedals<br /> * Amplifier<br /> * Cabinet<br /> * Microphone<br /> * Effects<br /> <br /> In Kemper terminology, &#039;&#039;&#039;Rig&#039;&#039;&#039; also has a specific platform meaning.<br /> <br /> === Rig Exchange ===<br /> <br /> Kemper&#039;s online service for sharing PROFILER Rigs.<br /> <br /> === RMS ===<br /> <br /> &#039;&#039;&#039;Root Mean Square.&#039;&#039;&#039;<br /> <br /> A method of describing signal magnitude.<br /> <br /> For a sine wave, RMS voltage is commonly used when calculating dBu during calibration.<br /> <br /> === RTNeural ===<br /> <br /> An open-source neural-network inference engine optimized for real-time audio.<br /> <br /> RTNeural is used by projects including AIDA-X and GuitarML software.<br /> <br /> RTNeural itself is not a capture-model format.<br /> <br /> == S ==<br /> <br /> === Sample rate ===<br /> <br /> The number of digital audio samples processed per second.<br /> <br /> Common rates include:<br /> <br /> * 44.1 kHz<br /> * 48 kHz<br /> * 88.2 kHz<br /> * 96 kHz<br /> <br /> Capture/training systems may require a specific sample rate.<br /> <br /> === Send ===<br /> <br /> The signal leaving the capture or recording system and traveling toward the device under test.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;interface send → reamp box → amplifier&#039;&#039;&#039;<br /> <br /> === Signal chain ===<br /> <br /> The ordered path through which an audio signal travels.<br /> <br /> The signal chain determines which equipment contributes to a capture.<br /> <br /> === Snapshot ===<br /> <br /> See [[#Snapshot model|Snapshot model]].<br /> <br /> === Snapshot model ===<br /> <br /> A model representing equipment at a particular configuration.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;amplifier gain = 5&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;amplifier gain = 8&#039;&#039;&#039;<br /> <br /> may require separate snapshot models.<br /> <br /> Snapshot modeling contrasts with conditioned modeling, in which one model can respond to a control parameter.<br /> <br /> === Stimulus ===<br /> <br /> See [[#Capture signal|Capture signal]].<br /> <br /> === Studio capture ===<br /> <br /> A capture of studio equipment such as:<br /> <br /> * Preamp<br /> * Saturation device<br /> * Channel strip<br /> * Compressor<br /> <br /> Whether a particular device can be captured accurately depends on the capabilities of the capture architecture.<br /> <br /> === System identification ===<br /> <br /> The engineering process of constructing a mathematical model of a system from observed input and output behavior.<br /> <br /> Capture and profiling technologies can be viewed as forms of nonlinear system identification.<br /> <br /> == T ==<br /> <br /> === Test signal ===<br /> <br /> See [[#Capture signal|Capture signal]].<br /> <br /> === Tone Model ===<br /> <br /> IK Multimedia&#039;s term for a model created for the [[TONEX]] ecosystem using AI Machine Modeling.<br /> <br /> A Tone Model is not a NAM model.<br /> <br /> === ToneNET ===<br /> <br /> IK Multimedia&#039;s online service for discovering and sharing TONEX Tone Models and related content.<br /> <br /> === Training ===<br /> <br /> The computational process that adjusts a model&#039;s parameters so its predicted output approximates the measured output of the physical device.<br /> <br /> Conceptually:<br /> <br /> &#039;&#039;&#039;known input + measured output → training → model&#039;&#039;&#039;<br /> <br /> Training should be distinguished from the physical recording or capture stage.<br /> <br /> === Training data ===<br /> <br /> Input/output measurements used to train a model.<br /> <br /> For capture technology this commonly consists of:<br /> <br /> * Known training input<br /> * Recorded response of the physical equipment<br /> <br /> === Training error ===<br /> <br /> A numerical measurement of the difference between model predictions and the target training data.<br /> <br /> Lower training error can indicate a closer mathematical fit, but it is not a complete measurement of perceived audio quality.<br /> <br /> === Training signal ===<br /> <br /> See [[#Capture signal|Capture signal]].<br /> <br /> == V ==<br /> <br /> === V1 ===<br /> <br /> The original Neural DSP Neural Capture generation.<br /> <br /> V1 training occurs locally on compatible Cortex hardware.<br /> <br /> See [[Neural Capture]].<br /> <br /> === V2 ===<br /> <br /> The second-generation Neural DSP Neural Capture technology.<br /> <br /> V2 uses higher-resolution cloud training and is designed to improve reproduction of difficult dynamic targets including fuzzes and compressors.<br /> <br /> See [[Neural Capture]].<br /> <br /> === VST3 ===<br /> <br /> A plugin format widely used by digital audio workstations.<br /> <br /> Several capture-model players are available as VST3 plugins.<br /> <br /> == W ==<br /> <br /> === WaveNet ===<br /> <br /> A convolutional neural-network architecture originally developed for audio generation.<br /> <br /> WaveNet-style architectures have also been adapted for nonlinear audio-system modeling.<br /> <br /> Some GuitarML projects have used WaveNet-based architectures.<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> * [[Neural Amp Modeler]]<br /> * [[NAM Model File Format]]<br /> * [[NAM A2 Architecture]]<br /> * [[NAM Playback]]<br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Neural Capture]]<br /> * [[Line 6 Proxy]]<br /> * [[Open Capture Technologies]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss capture technology and terminology on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Glossary]]<br /> [[Category:Capture technology]]<br /> [[Category:Reference]]</div> Tue, 01 Sep 2026 06:12:46 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Glossary https://namforum.com/wiki/index.php?title=Capture_Technologies&diff=21&oldid=0 https://namforum.com/wiki/index.php?title=Capture_Technologies&diff=21&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Capture Technologies =<br /> <br /> &#039;&#039;&#039;Capture technology&#039;&#039;&#039; is a general term for systems that create digital representations of physical audio equipment by measuring how that equipment responds to known input signals.<br /> <br /> Capture systems are commonly used to reproduce:<br /> <br /> * Guitar amplifiers<br /> * Bass amplifiers<br /> * Preamps<br /> * Overdrive pedals<br /> * Distortion pedals<br /> * Fuzz pedals<br /> * Compressors<br /> * Speaker/cabinet chains<br /> * Studio equipment<br /> * Complete signal chains<br /> <br /> Modern capture technologies include [[Neural Amp Modeler]], [[TONEX]], [[Kemper Profiling]], [[Neural Capture]], [[Line 6 Proxy]], and a variety of [[Open Capture Technologies|open capture technologies]].<br /> <br /> Although these systems pursue broadly similar goals, their model formats, training methods, playback engines, hardware requirements, and ecosystems are different.<br /> <br /> == Capture versus conventional modeling ==<br /> <br /> Traditional algorithmic modeling usually begins with a developer designing a mathematical or computational model intended to reproduce the behavior of a particular device or circuit.<br /> <br /> A capture system takes a different approach.<br /> <br /> It measures a physical device.<br /> <br /> Conceptually:<br /> <br /> &#039;&#039;&#039;known input → physical device → measured output&#039;&#039;&#039;<br /> <br /> A training or profiling process then determines a model that reproduces the relationship between the input and output.<br /> <br /> The resulting model can process new audio in real time.<br /> <br /> == What is being captured? ==<br /> <br /> A capture does not necessarily represent one individual piece of equipment.<br /> <br /> It represents the behavior of the entire measured signal path.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;test signal → amplifier → load box → recorder&#039;&#039;&#039;<br /> <br /> creates a different measurement from:<br /> <br /> &#039;&#039;&#039;test signal → amplifier → speaker → microphone → recorder&#039;&#039;&#039;<br /> <br /> The second path contains the characteristics of the:<br /> <br /> * Amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Return signal path<br /> <br /> This is why understanding the [[Capture Signal Chain]] is essential.<br /> <br /> == Snapshot captures ==<br /> <br /> Many capture systems create &#039;&#039;&#039;snapshot models&#039;&#039;&#039;.<br /> <br /> A snapshot represents equipment at a particular configuration.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Amplifier gain: 7&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Bass: 5&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Middle: 6&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Treble: 4&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Master: 3&#039;&#039;&#039;<br /> <br /> The resulting model represents the measured behavior around that configuration.<br /> <br /> Changing playback controls afterward does not necessarily reproduce the exact behavior that would have occurred if the physical amplifier controls had been changed.<br /> <br /> For substantially different physical settings, separate captures may be appropriate.<br /> <br /> == Conditioned models ==<br /> <br /> Some systems extend capture technology by including one or more control values as inputs to the model.<br /> <br /> This is known as &#039;&#039;&#039;conditioned modeling&#039;&#039;&#039;.<br /> <br /> Conceptually, a snapshot model learns:<br /> <br /> &#039;&#039;&#039;audio input → audio output&#039;&#039;&#039;<br /> <br /> A conditioned model can learn:<br /> <br /> &#039;&#039;&#039;audio input + control value → audio output&#039;&#039;&#039;<br /> <br /> This can allow one neural model to reproduce behavior across part of a physical control&#039;s range.<br /> <br /> Open-source projects such as GuitarML Proteus have experimented with conditioned models.<br /> <br /> Commercial systems may also combine captured behavior with separately modeled controls.<br /> <br /> == Hybrid capture and modeling ==<br /> <br /> The boundary between capture technology and conventional modeling is increasingly blurred.<br /> <br /> For example, Kemper&#039;s [[Kemper Profiling|Liquid Profiling]] combines a measured PROFILE with modeled gain and tone-stack behavior.<br /> <br /> A modern system may therefore contain:<br /> <br /> &#039;&#039;&#039;capture + algorithmic modeling + conventional effects&#039;&#039;&#039;<br /> <br /> rather than fitting entirely into one category.<br /> <br /> == Major capture platforms ==<br /> <br /> Several capture technologies have developed substantial user ecosystems.<br /> <br /> === Neural Amp Modeler ===<br /> <br /> [[Neural Amp Modeler]] is an open-source neural modeling system.<br /> <br /> NAM provides:<br /> <br /> * Open-source training<br /> * Open model specification<br /> * Open-source real-time playback core<br /> * Desktop software<br /> * Third-party hardware implementations<br /> * Community model sharing<br /> <br /> NAM models normally use the &#039;&#039;.nam&#039;&#039; file format.<br /> <br /> Current NAM snapshot models use the [[NAM A2 Architecture|A2 architecture]], while older A1 models remain supported.<br /> <br /> See [[NAM Playback]].<br /> <br /> === TONEX ===<br /> <br /> [[TONEX]] is IK Multimedia&#039;s proprietary AI Machine Modeling ecosystem.<br /> <br /> It provides:<br /> <br /> * TONEX Modeler<br /> * TONEX Player<br /> * TONEX Editor<br /> * ToneNET<br /> * TONEX Pedal<br /> * TONEX ONE<br /> * Other TONEX hardware<br /> <br /> Captured models are called &#039;&#039;&#039;Tone Models&#039;&#039;&#039;.<br /> <br /> The current TONEX Modeler separates physical capture from model training and can preserve capture data for later processing.<br /> <br /> === Kemper Profiling ===<br /> <br /> [[Kemper Profiling]] is the proprietary capture technology used by the KEMPER PROFILER ecosystem.<br /> <br /> Captured models are called &#039;&#039;&#039;PROFILEs&#039;&#039;&#039;.<br /> <br /> Kemper also provides:<br /> <br /> * Liquid Profiling<br /> * PROFILING 2.0<br /> * Rig Manager<br /> * Rig Exchange<br /> * Dedicated PROFILER hardware<br /> <br /> Liquid Profiling combines captured amplifier behavior with modeled control behavior.<br /> <br /> === Neural Capture ===<br /> <br /> [[Neural Capture]] is Neural DSP&#039;s proprietary capture technology for the Cortex ecosystem.<br /> <br /> Current systems support:<br /> <br /> * Neural Capture V1<br /> * Neural Capture V2<br /> <br /> V1 provides fast local capture.<br /> <br /> V2 uses higher-resolution cloud-based training and is intended to improve modeling of difficult dynamic equipment such as fuzzes, compressors, and highly responsive amplifiers.<br /> <br /> Neural Captures can be shared through Cortex Cloud.<br /> <br /> === Line 6 Proxy ===<br /> <br /> [[Line 6 Proxy]] is Line 6&#039;s capture technology for Helix Stadium.<br /> <br /> Captured models are called &#039;&#039;&#039;Clones&#039;&#039;&#039;.<br /> <br /> Current Proxy Clone types include:<br /> <br /> * Amp+Cab<br /> * Amp<br /> * Preamp<br /> * Distortion<br /> <br /> Helix Stadium performs the physical measurement and Line 6&#039;s online infrastructure performs model creation.<br /> <br /> Proxy Clones can be used in compatible Stadium hardware and software.<br /> <br /> === Open capture technologies ===<br /> <br /> Other open-source systems include projects such as:<br /> <br /> * AIDA-X<br /> * GuitarML Proteus<br /> * NeuralPi<br /> * SmartGuitarAmp<br /> <br /> See [[Open Capture Technologies]].<br /> <br /> These projects explore different neural architectures, model formats, training systems, and hardware targets.<br /> <br /> They should not automatically be assumed to be compatible with NAM or with one another.<br /> <br /> == Comparison ==<br /> <br /> The following table provides a broad conceptual comparison.<br /> <br /> {| class=&quot;wikitable&quot;<br /> ! Technology<br /> ! Model name<br /> ! Ecosystem<br /> ! User capture<br /> ! Training location<br /> ! Sharing<br /> |-<br /> | [[Neural Amp Modeler]]<br /> | NAM model<br /> | Open source<br /> | Yes<br /> | Local / cloud services<br /> | Independent / community services<br /> |-<br /> | [[TONEX]]<br /> | Tone Model<br /> | IK Multimedia<br /> | Yes<br /> | Computer<br /> | ToneNET<br /> |-<br /> | [[Kemper Profiling]]<br /> | PROFILE<br /> | Kemper<br /> | Yes<br /> | PROFILER hardware<br /> | Rig Exchange / independent<br /> |-<br /> | [[Neural Capture]]<br /> | Neural Capture<br /> | Neural DSP Cortex<br /> | Yes<br /> | Hardware for V1 / cloud for V2<br /> | Cortex Cloud<br /> |-<br /> | [[Line 6 Proxy]]<br /> | Clone<br /> | Line 6 Helix Stadium<br /> | Yes<br /> | Line 6 cloud<br /> | CustomTone / file exchange<br /> |-<br /> | [[Open Capture Technologies|AIDA-X / GuitarML / others]]<br /> | Varies<br /> | Open source<br /> | Varies<br /> | Usually computer / cloud workflow<br /> | Varies<br /> |}<br /> <br /> This table describes the broad architecture of each ecosystem rather than every possible workflow.<br /> <br /> == Model formats are not interchangeable ==<br /> <br /> One of the most important facts about capture technology is that there is currently no universal model format.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;NAM model ≠ TONEX Tone Model&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;TONEX Tone Model ≠ Kemper PROFILE&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Kemper PROFILE ≠ Neural Capture&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Neural Capture ≠ Proxy Clone&#039;&#039;&#039;<br /> <br /> Even open-source neural models are not necessarily interchangeable.<br /> <br /> Each system may use a different:<br /> <br /> * Neural architecture<br /> * Model representation<br /> * File format<br /> * Input normalization<br /> * Sample rate<br /> * Metadata<br /> * Playback engine<br /> <br /> A model must be interpreted by software that understands its architecture and format.<br /> <br /> == Capture versus training ==<br /> <br /> The terms &#039;&#039;&#039;capture&#039;&#039;&#039; and &#039;&#039;&#039;training&#039;&#039;&#039; are sometimes used as though they mean the same thing.<br /> <br /> They are conceptually different.<br /> <br /> &#039;&#039;&#039;Capture&#039;&#039;&#039; is the measurement of the physical equipment.<br /> <br /> &#039;&#039;&#039;Training&#039;&#039;&#039; is the computational process that creates a model from the measurement.<br /> <br /> A simplified workflow is:<br /> <br /> &#039;&#039;&#039;training signal → physical equipment → recorded response → training → model&#039;&#039;&#039;<br /> <br /> Some systems hide this distinction from the user.<br /> <br /> Others make it explicit.<br /> <br /> For example:<br /> <br /> * NAM training can occur after the physical recording.<br /> * TONEX Modeler can preserve capture data for later training.<br /> * Neural Capture V2 performs physical measurement on Cortex hardware and training in Cortex Cloud.<br /> * Proxy performs physical measurement with Stadium and model creation using Line 6&#039;s online infrastructure.<br /> <br /> Separating these concepts is useful when discussing capture technology accurately.<br /> <br /> == The capture recording ==<br /> <br /> The recording produced during the physical measurement is not itself the finished model.<br /> <br /> It is evidence of how the equipment responded to the training signal.<br /> <br /> Training uses that evidence to construct a model.<br /> <br /> Therefore:<br /> <br /> &#039;&#039;&#039;measurement ≠ model&#039;&#039;&#039;<br /> <br /> This distinction has important consequences.<br /> <br /> If the original measurement is preserved, it may sometimes be possible to train another model from it later.<br /> <br /> This can be useful when:<br /> <br /> * Training software improves<br /> * A new architecture appears<br /> * A different quality setting is desired<br /> * The original equipment is no longer available<br /> <br /> Whether old measurements can actually be reused depends on the requirements of the new training system.<br /> <br /> == Training signals ==<br /> <br /> Capture systems use carefully designed signals to stimulate the equipment being measured.<br /> <br /> A useful training signal must provide enough information for the system to learn the target&#039;s behavior.<br /> <br /> Different technologies use different training signals and procedures.<br /> <br /> They may contain:<br /> <br /> * Noise-like material<br /> * Sweeps<br /> * Impulses<br /> * Bursts<br /> * Guitar-like material<br /> * Specially designed excitation sequences<br /> <br /> The training signal should not normally be substituted with arbitrary audio unless the particular system explicitly supports it.<br /> <br /> == Nonlinear behavior ==<br /> <br /> Capture technologies are particularly useful for nonlinear audio equipment.<br /> <br /> Examples include:<br /> <br /> * Tube amplifiers<br /> * Solid-state amplifiers<br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> * Compression<br /> * Saturation<br /> <br /> A nonlinear device does not simply apply one fixed frequency response.<br /> <br /> Its output depends on the signal entering it.<br /> <br /> For example, a guitar amplifier may respond differently to:<br /> <br /> * Quiet picking<br /> * Hard picking<br /> * Low guitar volume<br /> * High guitar volume<br /> * Different input levels<br /> <br /> A successful capture must reproduce enough of this behavior to remain convincing with new playing material.<br /> <br /> == Linear systems ==<br /> <br /> Some parts of an audio chain are approximately linear and can be represented efficiently by an impulse response.<br /> <br /> The most common example is a:<br /> <br /> &#039;&#039;&#039;speaker + cabinet + microphone&#039;&#039;&#039;<br /> <br /> signal path.<br /> <br /> A conventional cabinet IR can reproduce its linear frequency and phase response without requiring a neural network.<br /> <br /> This is why many capture systems use:<br /> <br /> &#039;&#039;&#039;neural amplifier model → cabinet IR&#039;&#039;&#039;<br /> <br /> rather than requiring the neural model to reproduce the cabinet.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Direct captures ==<br /> <br /> A direct amplifier capture excludes the physical speaker and microphone response.<br /> <br /> A typical signal path is:<br /> <br /> &#039;&#039;&#039;training signal → amplifier → suitable load/direct output → recorder&#039;&#039;&#039;<br /> <br /> The resulting model normally requires cabinet processing when played through a full-range system.<br /> <br /> Advantages include:<br /> <br /> * Cabinet can be changed later<br /> * IRs can be exchanged<br /> * Physical guitar cabinets can be used<br /> * Amplifier and cabinet can be evaluated separately<br /> <br /> See [[Capture Types]].<br /> <br /> == Complete rig captures ==<br /> <br /> A complete rig capture can include:<br /> <br /> * Pedal<br /> * Amplifier<br /> * Cabinet<br /> * Speaker<br /> * Microphone<br /> * Microphone preamp<br /> * Other compatible processing<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;overdrive → amplifier → cabinet → microphone&#039;&#039;&#039;<br /> <br /> may be captured as one model.<br /> <br /> This can reproduce a specific finished sound very effectively.<br /> <br /> The tradeoff is flexibility.<br /> <br /> The individual components cannot necessarily be separated after they have been learned as one combined system.<br /> <br /> == Capturing pedals ==<br /> <br /> Capture technologies can model many gain-based pedals.<br /> <br /> Common targets include:<br /> <br /> * Boost<br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> <br /> Pedals can be captured:<br /> <br /> * Individually<br /> * With an amplifier<br /> * As part of a larger chain<br /> <br /> Capturing them individually provides greater routing flexibility.<br /> <br /> Capturing the combined chain can preserve interactions between the devices.<br /> <br /> == Difficult devices ==<br /> <br /> Not every audio processor is equally easy to capture.<br /> <br /> Potentially difficult targets include:<br /> <br /> * Fuzz circuits<br /> * Compressors<br /> * Devices with long-term state<br /> * Strongly dynamic circuits<br /> * Devices sensitive to source impedance<br /> * Time-varying effects<br /> <br /> A system that performs extremely well on an amplifier should not automatically be assumed to reproduce every compressor, fuzz, delay, or modulation effect equally well.<br /> <br /> The architecture and training method matter.<br /> <br /> == Time-based effects ==<br /> <br /> Snapshot neural capture systems generally focus on relatively short-memory nonlinear systems.<br /> <br /> Effects such as:<br /> <br /> * Delay<br /> * Reverb<br /> * Chorus<br /> * Flanger<br /> * Phaser<br /> * Tremolo<br /> <br /> may require different modeling approaches.<br /> <br /> Some capture systems explicitly instruct users not to include these effects during capture.<br /> <br /> When capturing a complete amplifier rig, disable time-based onboard effects unless the capture technology specifically supports them.<br /> <br /> == Input level ==<br /> <br /> Input level is fundamental to capture technology.<br /> <br /> For a nonlinear device:<br /> <br /> &#039;&#039;&#039;changing input level can change device behavior&#039;&#039;&#039;<br /> <br /> not merely volume.<br /> <br /> If the playback system drives a model differently from the original capture conditions, the result may exhibit different:<br /> <br /> * Gain<br /> * Distortion<br /> * Compression<br /> * Dynamics<br /> * Sustain<br /> * Cleanup<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Calibration ==<br /> <br /> Calibration attempts to preserve the relationship between digital signal level and the analog level used during capture.<br /> <br /> A calibrated workflow can improve model portability between different audio interfaces and playback systems.<br /> <br /> Not every capture platform exposes calibration in the same way.<br /> <br /> Some systems manage much of the process internally.<br /> <br /> Others provide explicit metadata or require the creator to document the signal chain.<br /> <br /> Calibration should therefore be considered part of the capture methodology rather than a universal feature implemented identically by every platform.<br /> <br /> == Reamping ==<br /> <br /> A capture system often needs to send a recorded training signal from an audio interface into equipment designed for an instrument-level source.<br /> <br /> A &#039;&#039;&#039;reamp device&#039;&#039;&#039; can provide appropriate:<br /> <br /> * Level conversion<br /> * Connection<br /> * Ground isolation<br /> * Signal interfacing<br /> <br /> The exact requirements depend on the equipment and interface.<br /> <br /> See [[Reamping for Capture]].<br /> <br /> == Accuracy ==<br /> <br /> Capture accuracy is not one simple number.<br /> <br /> A model can differ from the original in:<br /> <br /> * Frequency response<br /> * Harmonic distortion<br /> * Dynamic response<br /> * Transients<br /> * Compression<br /> * Sustain<br /> * Low-frequency behavior<br /> * High-frequency behavior<br /> * Response to guitar-volume changes<br /> <br /> A numerical training error can be useful, but it should not be treated as a complete substitute for listening.<br /> <br /> The final model should be evaluated using relevant musical material.<br /> <br /> == Level matching comparisons ==<br /> <br /> When comparing a capture against the original device, levels should be matched closely.<br /> <br /> Human hearing commonly interprets a slightly louder signal as:<br /> <br /> * Fuller<br /> * Clearer<br /> * More detailed<br /> * More exciting<br /> <br /> An unmatched comparison can therefore produce misleading conclusions.<br /> <br /> This principle also applies when comparing two different capture technologies.<br /> <br /> == Comparing capture systems ==<br /> <br /> A meaningful comparison should control as many variables as possible.<br /> <br /> Ideally use:<br /> <br /> * Same physical device<br /> * Same physical settings<br /> * Same source recording<br /> * Same cabinet/IR<br /> * Same playback level<br /> * Same monitoring system<br /> * Appropriate calibration<br /> * Equivalent capture type<br /> <br /> Otherwise the comparison may measure differences in the surrounding workflow rather than differences in the capture technology itself.<br /> <br /> == Model accuracy versus usability ==<br /> <br /> The mathematically closest model is not automatically the most useful model.<br /> <br /> Practical factors also include:<br /> <br /> * CPU usage<br /> * Latency<br /> * Training time<br /> * Hardware cost<br /> * Ease of capture<br /> * Model availability<br /> * Editing<br /> * Preset management<br /> * Live reliability<br /> * Sharing<br /> * Portability<br /> <br /> Different technologies make different engineering tradeoffs.<br /> <br /> == Native and converted models ==<br /> <br /> Some playback systems execute the original model architecture directly.<br /> <br /> Others import a model and convert or approximate it using another internal architecture.<br /> <br /> Conceptually:<br /> <br /> &#039;&#039;&#039;native&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;model → original-compatible engine → audio&#039;&#039;&#039;<br /> <br /> versus:<br /> <br /> &#039;&#039;&#039;conversion&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;model → conversion → different model architecture → audio&#039;&#039;&#039;<br /> <br /> Both approaches can be useful.<br /> <br /> They should nevertheless be described accurately.<br /> <br /> See [[NAM Playback]].<br /> <br /> == Hardware and software ==<br /> <br /> Capture models can be played using:<br /> <br /> * Desktop plugins<br /> * Standalone applications<br /> * Dedicated pedals<br /> * Multi-effects processors<br /> * Embedded systems<br /> * Mobile/portable hardware<br /> <br /> The model and player are separate concepts.<br /> <br /> A particular capture technology may support many players or only hardware from one manufacturer.<br /> <br /> == Open and proprietary ecosystems ==<br /> <br /> Capture technologies span a spectrum from open-source systems to vertically integrated proprietary products.<br /> <br /> An open ecosystem may publish:<br /> <br /> * Training code<br /> * Playback code<br /> * Model specification<br /> * Development tools<br /> <br /> A proprietary ecosystem may integrate:<br /> <br /> * Capture<br /> * Training<br /> * Playback<br /> * Hardware<br /> * Cloud services<br /> * Model sharing<br /> <br /> Neither structure automatically determines modeling quality.<br /> <br /> It does affect:<br /> <br /> * Interoperability<br /> * Third-party development<br /> * Long-term access<br /> * Hardware choices<br /> * Model distribution<br /> <br /> == Community libraries ==<br /> <br /> Capture technology has created large libraries of user-generated models.<br /> <br /> Examples include services associated with:<br /> <br /> * NAM<br /> * TONEX<br /> * Kemper<br /> * Neural DSP<br /> * Line 6<br /> <br /> Community sharing dramatically increases the usefulness of capture systems.<br /> <br /> It also creates documentation problems.<br /> <br /> A model without information about what was captured can be difficult to evaluate or reuse.<br /> <br /> == Model documentation ==<br /> <br /> Useful capture documentation includes:<br /> <br /> * Platform<br /> * Model architecture/version<br /> * Equipment manufacturer<br /> * Equipment model<br /> * Equipment settings<br /> * Capture type<br /> * Cabinet<br /> * Speaker<br /> * Microphone<br /> * Microphone position<br /> * Load box<br /> * Calibration<br /> * Capture date<br /> * Creator<br /> * Training software/version<br /> * Notes<br /> <br /> Not every field applies to every capture.<br /> <br /> The objective is to preserve enough information to understand what the model actually represents.<br /> <br /> == Model names are not documentation ==<br /> <br /> A filename such as:<br /> <br /> &#039;&#039;&#039;Marshall High Gain&#039;&#039;&#039;<br /> <br /> does not reveal whether the model contains:<br /> <br /> * Amplifier only<br /> * Amplifier and cabinet<br /> * Overdrive and amplifier<br /> * Microphone<br /> * Particular channel<br /> * Particular amplifier settings<br /> <br /> Do not infer the capture signal chain solely from a model name.<br /> <br /> See [[Capture Types]].<br /> <br /> == Reproducibility ==<br /> <br /> A well-documented capture can potentially be recreated or investigated later.<br /> <br /> Useful records include:<br /> <br /> * Original training signal<br /> * Original returned recording<br /> * Equipment settings<br /> * Photographs<br /> * Signal-chain diagram<br /> * Calibration data<br /> * Software versions<br /> * Finished model<br /> <br /> Reproducibility becomes particularly valuable when comparing training algorithms or architectures.<br /> <br /> == Preservation ==<br /> <br /> Capture models may outlive the software or hardware originally used to create them.<br /> <br /> For important captures, consider preserving:<br /> <br /> * Finished model files<br /> * Original measurements<br /> * Metadata<br /> * Documentation<br /> * Required player versions<br /> * Relevant open-source software where legally appropriate<br /> <br /> This is especially important for unique equipment that may later be sold, modified, damaged, or unavailable.<br /> <br /> == Capture as documentation of equipment ==<br /> <br /> Capture technology has uses beyond replacing physical equipment.<br /> <br /> A well-documented model can preserve aspects of:<br /> <br /> * Rare amplifiers<br /> * Modified equipment<br /> * Unique pedals<br /> * Studio signal chains<br /> * Artist rigs<br /> * Historical equipment<br /> <br /> A capture is not a complete physical simulation of the original device, but it can preserve a useful measurement of how that device behaved at a particular time and configuration.<br /> <br /> == Choosing a capture technology ==<br /> <br /> The best system depends on the intended use.<br /> <br /> Consider:<br /> <br /> * Equipment to be captured<br /> * Desired accuracy<br /> * Hardware already owned<br /> * Live requirements<br /> * DAW requirements<br /> * CPU resources<br /> * Training workflow<br /> * Cloud requirements<br /> * Open-source requirements<br /> * Model-sharing community<br /> * Need for conditioned controls<br /> * Cabinet workflow<br /> * Budget<br /> <br /> There is no single capture platform that is automatically best for every user or every device.<br /> <br /> == NAMFORUM scope ==<br /> <br /> NAMFORUM covers capture technology as a field rather than limiting discussion to one platform.<br /> <br /> This includes:<br /> <br /> * [[Neural Amp Modeler]]<br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Neural Capture]]<br /> * [[Line 6 Proxy]]<br /> * [[Open Capture Technologies|Open capture systems]]<br /> * Emerging capture technologies<br /> * Capture hardware<br /> * Training<br /> * Calibration<br /> * Reamping<br /> * Model sharing<br /> * Playback<br /> * Research<br /> <br /> Conventional algorithmic modeling may also be discussed where it intersects with capture technology, but capture and profiling technologies are the primary focus.<br /> <br /> == See also ==<br /> <br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> * [[Neural Amp Modeler]]<br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Neural Capture]]<br /> * [[Line 6 Proxy]]<br /> * [[Open Capture Technologies]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss capture, profiling, training, and modeling technology on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Audio technology]]</div> Tue, 01 Sep 2026 06:10:10 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Capture_Technologies https://namforum.com/wiki/index.php?title=Open_Capture_Technologies&diff=20&oldid=0 https://namforum.com/wiki/index.php?title=Open_Capture_Technologies&diff=20&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Open Capture Technologies =<br /> <br /> &#039;&#039;&#039;Open capture technologies&#039;&#039;&#039; are open-source projects that use machine learning or related techniques to reproduce the behavior of amplifiers, pedals, preamps, and other audio equipment.<br /> <br /> [[Neural Amp Modeler]] is currently one of the best-known open capture ecosystems, but it is not the only open-source project in this field.<br /> <br /> Other projects have explored different:<br /> <br /> * Neural-network architectures<br /> * Training methods<br /> * Model formats<br /> * Playback engines<br /> * Plugin formats<br /> * Embedded hardware<br /> * Conditioned models<br /> * Snapshot models<br /> <br /> Important examples include &#039;&#039;&#039;AIDA-X&#039;&#039;&#039; and projects developed by &#039;&#039;&#039;GuitarML&#039;&#039;&#039;, including &#039;&#039;&#039;Proteus&#039;&#039;&#039;, &#039;&#039;&#039;NeuralPi&#039;&#039;&#039;, and earlier SmartGuitarAmp/SmartPedal work.<br /> <br /> These technologies are related conceptually but should not be assumed to be model-compatible.<br /> <br /> == Open source does not mean one format ==<br /> <br /> There is no universal open-source capture-model format.<br /> <br /> An open-source project may define its own:<br /> <br /> * Model architecture<br /> * File format<br /> * Training procedure<br /> * Sample rate<br /> * Input normalization<br /> * Runtime engine<br /> * Metadata<br /> * Playback requirements<br /> <br /> Consequently:<br /> <br /> &#039;&#039;&#039;open-source model ≠ NAM model&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;neural amp model ≠ .nam file&#039;&#039;&#039;<br /> <br /> A model created for one project generally cannot be loaded directly by another unless explicit compatibility has been implemented.<br /> <br /> == Why open capture technologies matter ==<br /> <br /> Open implementations allow researchers, developers, musicians, and hardware manufacturers to examine and modify the underlying technology.<br /> <br /> Depending on the project&#039;s license and architecture, developers may be able to:<br /> <br /> * Study the training process<br /> * Study the playback implementation<br /> * Build alternative players<br /> * Create hardware implementations<br /> * Modify the neural architecture<br /> * Develop new training tools<br /> * Experiment with model conditioning<br /> * Improve performance<br /> * Create independent model libraries<br /> <br /> This differs from proprietary capture systems whose internal model representation and playback technology are controlled by one manufacturer.<br /> <br /> See [[Capture Technologies]].<br /> <br /> == AIDA-X ==<br /> <br /> &#039;&#039;&#039;AIDA-X&#039;&#039;&#039; is an open-source neural Amp Model player developed by Aida DSP.<br /> <br /> Its primary purpose is high-fidelity modeling of guitar amplifiers, although models can represent larger signal chains containing combinations of:<br /> <br /> * Amplifier<br /> * Cabinet<br /> * Distortion<br /> * Overdrive<br /> * Fuzz<br /> * Boost<br /> * EQ<br /> <br /> AIDA-X uses neural-network models trained from recorded input/output examples.<br /> <br /> The project provides both desktop/plugin playback and implementations intended for embedded hardware.<br /> <br /> == AIDA-X model playback ==<br /> <br /> AIDA-X uses &#039;&#039;&#039;RTNeural&#039;&#039;&#039; as its neural-network inference engine.<br /> <br /> The desktop implementation supports:<br /> <br /> * CLAP<br /> * LV2<br /> * VST2<br /> * VST3<br /> * Standalone operation<br /> <br /> Installers or binaries are available for Windows, macOS, and Linux.<br /> <br /> The project also maintains a headless LV2 implementation intended for embedded environments.<br /> <br /> Examples include systems based on:<br /> <br /> * MOD hardware<br /> * Raspberry Pi<br /> * Portenta X8<br /> * Aida DSP OS<br /> <br /> This makes AIDA-X notable not only as a desktop neural amplifier project but also as an open platform for embedded neural-model playback.<br /> <br /> == AIDA-X cabinet processing ==<br /> <br /> AIDA-X includes cabinet impulse-response processing after the Amp Model.<br /> <br /> Users can therefore combine a direct amplifier model with a cabinet IR within the player.<br /> <br /> Supported IR audio formats include WAV and FLAC.<br /> <br /> This follows the common modular signal-chain approach:<br /> <br /> &#039;&#039;&#039;neural amplifier model → cabinet IR&#039;&#039;&#039;<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Training AIDA-X models ==<br /> <br /> The AIDA-X project provides a model-training workflow.<br /> <br /> As with other capture systems, the general process involves:<br /> <br /> # Obtain or generate the required training input.<br /> # Send the signal through the equipment being modeled.<br /> # Record the resulting output.<br /> # Train the neural model from the corresponding input and output.<br /> # Load the finished model into AIDA-X.<br /> # Compare the model with the original equipment.<br /> <br /> The exact current training procedure should be obtained from the project&#039;s current trainer and documentation because open-source training tools can evolve independently of the playback application.<br /> <br /> == AIDA-X and NAM ==<br /> <br /> AIDA-X and [[Neural Amp Modeler]] are separate projects.<br /> <br /> Although both use neural networks to reproduce audio equipment, an AIDA-X model should not automatically be assumed to be a &#039;&#039;.nam&#039;&#039; model.<br /> <br /> The projects have their own:<br /> <br /> * Model architectures<br /> * Training systems<br /> * Playback implementations<br /> * Model formats<br /> <br /> The existence of RTNeural in one project and neural-network processing in another does not make their models interchangeable.<br /> <br /> Compatibility should always be verified explicitly.<br /> <br /> == GuitarML ==<br /> <br /> &#039;&#039;&#039;GuitarML&#039;&#039;&#039; is an open-source collection of machine-learning projects focused on guitar amplifiers, pedals, and related audio processing.<br /> <br /> GuitarML has produced several experimental and practical neural-modeling systems, including:<br /> <br /> * Proteus<br /> * NeuralPi<br /> * SmartGuitarAmp<br /> * SmartPedal<br /> * SmartAmp-related projects<br /> * Training utilities and model libraries<br /> <br /> These projects have explored multiple neural-network architectures rather than defining one universal GuitarML model format.<br /> <br /> == GuitarML Proteus ==<br /> <br /> &#039;&#039;&#039;Proteus&#039;&#039;&#039; is an open-source GuitarML plugin for capturing and playing models of amplifiers, pedals, plugins, and complete rigs.<br /> <br /> The project describes its purpose as providing functionality conceptually similar to systems such as Kemper, Quad Cortex, and TONEX through a free and open-source plugin.<br /> <br /> Proteus can create:<br /> <br /> * Snapshot models<br /> * Conditioned knob models<br /> <br /> This distinction makes Proteus particularly interesting within the history of open capture technology.<br /> <br /> == Proteus snapshot models ==<br /> <br /> A Proteus snapshot represents equipment at a particular operating point.<br /> <br /> Examples include:<br /> <br /> * Amplifier at one gain setting<br /> * Overdrive at one drive setting<br /> * Pedal/amplifier combination<br /> * Complete gain-based rig<br /> <br /> This is conceptually similar to snapshot capture in several other systems.<br /> <br /> The physical controls are configured before the training recording is made.<br /> <br /> == Proteus conditioned models ==<br /> <br /> Proteus can also capture the behavior of a variable control.<br /> <br /> The project refers to these as &#039;&#039;&#039;Knob&#039;&#039;&#039; models.<br /> <br /> Instead of learning only one fixed operating point, training data is recorded at multiple control positions.<br /> <br /> The resulting model can then respond to a parameter representing that control.<br /> <br /> This is an important distinction between:<br /> <br /> &#039;&#039;&#039;snapshot modeling&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;conditioned modeling&#039;&#039;&#039;<br /> <br /> A snapshot learns:<br /> <br /> &#039;&#039;&#039;audio input → audio output&#039;&#039;&#039;<br /> <br /> at a particular equipment configuration.<br /> <br /> A conditioned model can conceptually learn:<br /> <br /> &#039;&#039;&#039;audio input + control value → audio output&#039;&#039;&#039;<br /> <br /> allowing one neural model to represent part of the equipment&#039;s behavior across a control range.<br /> <br /> == Proteus neural architecture ==<br /> <br /> Proteus uses an &#039;&#039;&#039;LSTM&#039;&#039;&#039; neural network.<br /> <br /> The current project documentation specifies an LSTM hidden size of 40 for compatible Proteus models.<br /> <br /> The real-time neural-network processing is performed using RTNeural.<br /> <br /> Proteus model files use JSON.<br /> <br /> These files are specific to the Proteus architecture and should not be confused with NAM files merely because both formats can use JSON internally.<br /> <br /> == Capturing with Proteus ==<br /> <br /> The Proteus Capture Utility provides the input audio and training workflow required to create models.<br /> <br /> A typical capture process is:<br /> <br /> # Play the Proteus capture WAV through the equipment.<br /> # Record the equipment output.<br /> # Export the recording.<br /> # Upload the recording to the provided Google Colab training workflow.<br /> # Train the model.<br /> # Download the resulting JSON model.<br /> # Load the model into Proteus.<br /> <br /> The project recommends using a reamp device where appropriate and a load box for direct amplifier captures.<br /> <br /> See [[Reamping for Capture]].<br /> <br /> == Proteus capture file requirements ==<br /> <br /> The current Proteus documentation specifies capture recordings as:<br /> <br /> * WAV<br /> * Mono<br /> * 44.1 kHz<br /> * PCM16 or FP32<br /> <br /> Timing alignment is particularly important.<br /> <br /> The project warns that latency mismatch between the training input and recorded output can prevent successful training.<br /> <br /> An initial click in the supplied capture signal can be used as an alignment reference.<br /> <br /> This illustrates a principle shared by many capture technologies:<br /> <br /> &#039;&#039;&#039;The training system must know which output samples correspond to which input samples.&#039;&#039;&#039;<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Proteus model targets ==<br /> <br /> Proteus is intended primarily for nonlinear equipment such as:<br /> <br /> * Amplifiers<br /> * Preamps<br /> * Distortion pedals<br /> * Overdrive pedals<br /> * Boost pedals<br /> <br /> The project explicitly advises against attempting to capture time-based effects such as:<br /> <br /> * Reverb<br /> * Delay<br /> * Flanger<br /> * Phaser<br /> <br /> A larger gain-based signal chain can also be captured as one model.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;overdrive → amplifier&#039;&#039;&#039;<br /> <br /> can be learned as one combined system.<br /> <br /> == Proteus direct amplifier models ==<br /> <br /> A direct amplifier model excludes the physical cabinet and microphone.<br /> <br /> Proteus recommends combining these models with an impulse response during playback.<br /> <br /> The signal chain becomes:<br /> <br /> &#039;&#039;&#039;Proteus amp model → cabinet IR&#039;&#039;&#039;<br /> <br /> Proteus includes a default IR for convenience, although the project recommends using a dedicated IR solution where appropriate.<br /> <br /> == Proteus Amp+Cab models ==<br /> <br /> An amplifier can also be captured using a microphone.<br /> <br /> In that case, the resulting model includes the measured:<br /> <br /> * Amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> <br /> An additional cabinet IR is normally unnecessary unless deliberately used for creative processing.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Proteus plugin formats ==<br /> <br /> Proteus is available in common desktop plugin formats.<br /> <br /> Current project releases provide formats including:<br /> <br /> * VST3<br /> * AU<br /> * AAX<br /> * LV2<br /> <br /> Platform availability varies by operating system.<br /> <br /> Because the project is open source, the code can also be compiled independently.<br /> <br /> == GuitarML NeuralPi ==<br /> <br /> &#039;&#039;&#039;NeuralPi&#039;&#039;&#039; is another GuitarML neural-modeling project.<br /> <br /> It was developed partly around running neural amplifier models on compact and embedded systems, including Raspberry Pi-related applications.<br /> <br /> NeuralPi supports both:<br /> <br /> * Snapshot models<br /> * Conditioned models<br /> <br /> Later versions added the ability to load models conditioned on a Gain/Drive parameter.<br /> <br /> This represents another exploration of variable-control neural modeling.<br /> <br /> == GuitarML model compatibility ==<br /> <br /> Even models created by different GuitarML projects are not necessarily interchangeable.<br /> <br /> The GuitarML ToneLibrary identifies different neural architectures for its projects.<br /> <br /> Examples include:<br /> <br /> &#039;&#039;&#039;Proteus&#039;&#039;&#039;<br /> <br /> Stateful LSTM followed by a dense layer.<br /> <br /> &#039;&#039;&#039;NeuralPi&#039;&#039;&#039;<br /> <br /> Stateful LSTM followed by a dense layer.<br /> <br /> &#039;&#039;&#039;SmartAmpPro&#039;&#039;&#039;<br /> <br /> Convolutional layers followed by a stateless LSTM and dense layer.<br /> <br /> &#039;&#039;&#039;SmartPedal / SmartAmp&#039;&#039;&#039;<br /> <br /> WaveNet.<br /> <br /> Because the architectures differ, models should be loaded only into the software for which they were created unless explicit conversion or compatibility exists.<br /> <br /> == GuitarML ToneLibrary ==<br /> <br /> GuitarML maintains an open ToneLibrary containing model files for its projects.<br /> <br /> The library organizes models according to the intended player.<br /> <br /> This is necessary because a JSON model for one GuitarML architecture may not work with another GuitarML application.<br /> <br /> File extension alone is therefore not sufficient to determine compatibility.<br /> <br /> == SmartGuitarAmp ==<br /> <br /> &#039;&#039;&#039;SmartGuitarAmp&#039;&#039;&#039; was an earlier GuitarML project using a WaveNet neural network to reproduce amplifier behavior.<br /> <br /> The project helped demonstrate practical real-time neural guitar processing in an open-source plugin.<br /> <br /> Later GuitarML projects explored other architectures and more flexible model-loading and capture workflows.<br /> <br /> SmartGuitarAmp is historically important as part of the development path that led to the broader GuitarML ecosystem.<br /> <br /> == RTNeural ==<br /> <br /> &#039;&#039;&#039;RTNeural&#039;&#039;&#039; is a high-performance neural-network inference engine designed for real-time audio applications.<br /> <br /> It is used by projects including:<br /> <br /> * AIDA-X<br /> * GuitarML Proteus<br /> <br /> RTNeural is not itself an amplifier capture format.<br /> <br /> It provides infrastructure for efficiently executing supported neural-network architectures.<br /> <br /> Different applications can use RTNeural while still having completely incompatible model formats and architectures.<br /> <br /> This distinction is important.<br /> <br /> A common runtime technology does not imply a common model standard.<br /> <br /> == Snapshot versus conditioned modeling ==<br /> <br /> Open projects have been particularly useful for experimenting with the distinction between snapshot and conditioned models.<br /> <br /> A &#039;&#039;&#039;snapshot model&#039;&#039;&#039; represents one equipment configuration.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Amp gain = 6&#039;&#039;&#039;<br /> <br /> A separate capture might represent:<br /> <br /> &#039;&#039;&#039;Amp gain = 8&#039;&#039;&#039;<br /> <br /> A &#039;&#039;&#039;conditioned model&#039;&#039;&#039; can instead accept a control parameter as part of the neural-network input.<br /> <br /> Conceptually:<br /> <br /> &#039;&#039;&#039;audio + gain setting → modeled audio&#039;&#039;&#039;<br /> <br /> This allows one trained model to reproduce behavior across multiple settings.<br /> <br /> Conditioned modeling can potentially reduce the need for large collections of individual snapshots.<br /> <br /> However, training and accurately reproducing a variable control can be more complex than modeling one fixed operating point.<br /> <br /> == Capture technology is broader than products ==<br /> <br /> Commercial capture systems are often discussed as products:<br /> <br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Neural Capture]]<br /> * [[Line 6 Proxy]]<br /> <br /> Open-source capture technology is better understood as an evolving collection of:<br /> <br /> * Research<br /> * Code<br /> * Training systems<br /> * Runtime libraries<br /> * File formats<br /> * Plugins<br /> * Hardware implementations<br /> <br /> A research project can influence later products even if the original software itself never becomes a widely used commercial platform.<br /> <br /> == Research origins ==<br /> <br /> Modern neural amplifier modeling builds on academic research into real-time nonlinear audio-system modeling.<br /> <br /> Important approaches have included:<br /> <br /> * Recurrent neural networks<br /> * LSTM networks<br /> * WaveNet-style convolutional networks<br /> * Conditioned neural networks<br /> * Other temporal neural architectures<br /> <br /> Open-source projects have made many of these techniques accessible to musicians and developers without requiring proprietary hardware.<br /> <br /> == Model files are not audio recordings ==<br /> <br /> A neural model file is not simply the captured output recording.<br /> <br /> The recording is training data.<br /> <br /> Training analyzes the relationship between:<br /> <br /> &#039;&#039;&#039;known input&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;recorded output&#039;&#039;&#039;<br /> <br /> and produces a neural-network model.<br /> <br /> The model contains parameters that allow a compatible player to predict the equipment&#039;s response to new audio.<br /> <br /> Therefore:<br /> <br /> &#039;&#039;&#039;capture recording ≠ trained model&#039;&#039;&#039;<br /> <br /> Preserving both can be valuable.<br /> <br /> == Preserving training data ==<br /> <br /> Where the training system permits it, creators should preserve:<br /> <br /> * Original training input<br /> * Original recorded output<br /> * Equipment settings<br /> * Interface settings<br /> * Calibration information<br /> * Training software/version<br /> * Finished model<br /> * Notes<br /> <br /> The original measurement may remain useful even if the model architecture or training software changes later.<br /> <br /> A new neural architecture can potentially be trained from old measurement data if the new system accepts compatible training material.<br /> <br /> == Open formats and longevity ==<br /> <br /> One potential advantage of open capture technologies is long-term inspectability.<br /> <br /> If the:<br /> <br /> * Model specification<br /> * Playback code<br /> * Training code<br /> <br /> remain publicly available, models may remain usable even if the original developer stops maintaining a particular application.<br /> <br /> This does not guarantee permanent compatibility.<br /> <br /> Open-source projects can still change:<br /> <br /> * Architecture<br /> * File format<br /> * Dependencies<br /> * Operating-system support<br /> * Build systems<br /> <br /> Archiving the relevant source code and documentation can therefore be as important as preserving model files.<br /> <br /> == Open source and commercial use ==<br /> <br /> &quot;Open source&quot; does not necessarily mean that code can be incorporated into any commercial product without restrictions.<br /> <br /> Each project has a software license.<br /> <br /> Examples include licenses such as:<br /> <br /> * GPL<br /> * MIT<br /> * BSD<br /> * Other open-source licenses<br /> <br /> The rights and obligations differ substantially.<br /> <br /> Developers considering integration into hardware or commercial software should examine the actual license of every relevant component.<br /> <br /> Do not assume that freely downloadable source code is equivalent to unrestricted commercial licensing.<br /> <br /> == Comparing open capture systems ==<br /> <br /> Useful comparison criteria include:<br /> <br /> * Modeling accuracy<br /> * Neural architecture<br /> * Snapshot support<br /> * Conditioned-model support<br /> * Training time<br /> * Training hardware<br /> * CPU requirements<br /> * Latency<br /> * Model file format<br /> * Sample rate<br /> * Calibration<br /> * Metadata<br /> * Cabinet handling<br /> * Embedded support<br /> * Plugin support<br /> * Licensing<br /> * Model sharing<br /> * Development activity<br /> <br /> No single criterion determines whether a technology is useful.<br /> <br /> A small neural architecture may be ideal for embedded hardware even if a larger model achieves lower numerical error on a desktop computer.<br /> <br /> == Interoperability ==<br /> <br /> Open source makes interoperability possible, but it does not make interoperability automatic.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;AIDA-X model → Proteus&#039;&#039;&#039;<br /> <br /> is not automatically supported.<br /> <br /> &#039;&#039;&#039;Proteus model → NAM&#039;&#039;&#039;<br /> <br /> is not automatically supported.<br /> <br /> &#039;&#039;&#039;NAM model → AIDA-X&#039;&#039;&#039;<br /> <br /> is not automatically supported.<br /> <br /> A converter or compatible playback implementation must understand the source architecture and represent it correctly.<br /> <br /> Simply renaming a file extension does not convert a neural model.<br /> <br /> == Why multiple architectures exist ==<br /> <br /> Different neural architectures represent different engineering tradeoffs.<br /> <br /> Developers may optimize for:<br /> <br /> * Accuracy<br /> * CPU usage<br /> * Memory<br /> * Training speed<br /> * Embedded hardware<br /> * Parameter conditioning<br /> * Development simplicity<br /> * Real-time performance<br /> <br /> The existence of multiple open projects is therefore useful.<br /> <br /> They provide different approaches to the same broad problem rather than requiring one implementation to satisfy every use case.<br /> <br /> == Relationship to NAM ==<br /> <br /> [[Neural Amp Modeler]] is part of this broader open-source neural audio field.<br /> <br /> NAM has developed its own:<br /> <br /> * Model specification<br /> * Training system<br /> * Neural architectures<br /> * Real-time core<br /> * Plugin/player ecosystem<br /> * Hardware integrations<br /> * Model-sharing ecosystem<br /> <br /> Other open projects should not merely be described as NAM alternatives.<br /> <br /> They represent independent lines of development that have contributed to the larger field of neural audio modeling.<br /> <br /> == Documentation ==<br /> <br /> When sharing a model from an open capture system, identify at minimum:<br /> <br /> * Project/platform<br /> * Compatible player<br /> * Model architecture where relevant<br /> * Equipment<br /> * Capture type<br /> * Equipment settings<br /> * Cabinet status<br /> * Sample rate where relevant<br /> * Training method/version<br /> * Creator<br /> * Notes<br /> <br /> This is particularly important when several systems use generic extensions such as &#039;&#039;.json&#039;&#039;.<br /> <br /> A filename alone may provide no reliable indication of model compatibility.<br /> <br /> == Troubleshooting ==<br /> <br /> If an open-source model will not load, first determine:<br /> <br /> # Which project created the model.<br /> # Which architecture it uses.<br /> # Which player it was intended for.<br /> # Whether the player version supports that architecture.<br /> # Whether the model file is complete and valid.<br /> <br /> If the model loads but sounds incorrect, check:<br /> <br /> * Input level<br /> * Sample rate<br /> * Cabinet processing<br /> * IR configuration<br /> * Calibration<br /> * Output level<br /> * Additional processing<br /> * Training quality<br /> <br /> Do not assume that two neural-model players will produce equivalent results merely because both are open source.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Official resources ==<br /> <br /> * [https://github.com/AidaDSP/AIDA-X AIDA-X]<br /> * [https://github.com/GuitarML/Proteus GuitarML Proteus]<br /> * [https://github.com/GuitarML/NeuralPi GuitarML NeuralPi]<br /> * [https://github.com/GuitarML/ToneLibrary GuitarML ToneLibrary]<br /> * [https://github.com/GuitarML/SmartGuitarAmp SmartGuitarAmp]<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Neural Amp Modeler]]<br /> * [[NAM Model File Format]]<br /> * [[Training a NAM Model]]<br /> * [[Capture Types]]<br /> * [[Creating a Capture]]<br /> * [[Capture Signal Chain]]<br /> * [[Reamping for Capture]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss AIDA-X, GuitarML, Proteus, and other open capture technologies on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Open capture technologies]]<br /> [[Category:Capture technology]]<br /> [[Category:Neural audio]]</div> Tue, 01 Sep 2026 06:08:45 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Open_Capture_Technologies https://namforum.com/wiki/index.php?title=Line_6_Proxy&diff=19&oldid=0 https://namforum.com/wiki/index.php?title=Line_6_Proxy&diff=19&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Line 6 Proxy =<br /> <br /> &#039;&#039;&#039;Proxy&#039;&#039;&#039; is Line 6&#039;s capture and cloning technology for the &#039;&#039;&#039;Helix Stadium&#039;&#039;&#039; platform.<br /> <br /> Proxy allows users to measure physical amplifiers, preamps, speaker/cabinet setups, and gain-based pedals and create digital models called &#039;&#039;&#039;Clones&#039;&#039;&#039;.<br /> <br /> Proxy was introduced with &#039;&#039;&#039;Helix Stadium firmware 1.3&#039;&#039;&#039; in March 2026.<br /> <br /> Line 6 describes the initial implementation as &#039;&#039;&#039;Phase 1&#039;&#039;&#039; of the Proxy cloning engine.<br /> <br /> == Overview ==<br /> <br /> Proxy combines the audio hardware and routing capabilities of Helix Stadium with cloud-based model creation.<br /> <br /> A simplified workflow is:<br /> <br /> &#039;&#039;&#039;Helix Stadium test signal → physical equipment → Helix Stadium return → Line 6 cloud processing → Proxy Clone&#039;&#039;&#039;<br /> <br /> The Stadium hardware sends test signals through the equipment and records the resulting response.<br /> <br /> The collected training data is then uploaded to Line 6 servers, where the computationally intensive model creation takes place.<br /> <br /> The finished Clone is returned to the Stadium and can be saved in its Clone Library.<br /> <br /> == Supported Clone types ==<br /> <br /> The current Proxy implementation supports four primary Clone types:<br /> <br /> * &#039;&#039;&#039;Amp+Cab&#039;&#039;&#039;<br /> * &#039;&#039;&#039;Amp&#039;&#039;&#039;<br /> * &#039;&#039;&#039;Preamp&#039;&#039;&#039;<br /> * &#039;&#039;&#039;Distortion&#039;&#039;&#039;<br /> <br /> Each represents a different portion of the physical signal chain.<br /> <br /> == Amp+Cab Clone ==<br /> <br /> An &#039;&#039;&#039;Amp+Cab&#039;&#039;&#039; Clone captures an amplifier together with a physical speaker cabinet and microphone.<br /> <br /> A typical signal path is:<br /> <br /> &#039;&#039;&#039;Stadium → amplifier → speaker cabinet → microphone → Stadium&#039;&#039;&#039;<br /> <br /> Line 6 states that all elements in this path become part of the Clone, including:<br /> <br /> * Amplifier preamp<br /> * Power amplifier<br /> * Amplifier settings<br /> * Cabinet<br /> * Speaker<br /> * Microphone<br /> * Microphone position<br /> * Microphone distance<br /> <br /> This type of Clone therefore represents a complete miked amplifier signal chain.<br /> <br /> An additional conventional cabinet IR is normally unnecessary during playback unless deliberately used as an additional creative effect.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Amp Clone ==<br /> <br /> An &#039;&#039;&#039;Amp&#039;&#039;&#039; Clone captures an amplifier without the physical speaker cabinet.<br /> <br /> This requires a suitable load box with a direct output.<br /> <br /> A simplified path is:<br /> <br /> &#039;&#039;&#039;Stadium → amplifier → load box/direct output → Stadium&#039;&#039;&#039;<br /> <br /> The resulting Clone can then be combined with:<br /> <br /> * A Helix Stadium Cab block<br /> * A cabinet IR<br /> * Another compatible cabinet solution<br /> * A physical guitar cabinet with appropriate power amplification<br /> <br /> This provides more flexibility than an Amp+Cab Clone because the cabinet portion of the signal chain remains separate.<br /> <br /> == Amplifier safety ==<br /> <br /> &#039;&#039;&#039;Never connect an amplifier&#039;s speaker output directly to an ordinary Helix Stadium return input.&#039;&#039;&#039;<br /> <br /> An amplifier designed to operate into a speaker load must remain connected to an appropriate load.<br /> <br /> For direct Amp cloning, use a load box or other equipment specifically designed to accept the amplifier&#039;s speaker-level output and provide an appropriate direct signal.<br /> <br /> Incorrect connection can damage the amplifier, Stadium, or both.<br /> <br /> == Preamp Clone ==<br /> <br /> A &#039;&#039;&#039;Preamp&#039;&#039;&#039; Clone represents a standalone preamp or the preamp section of an amplifier.<br /> <br /> For a standalone preamp:<br /> <br /> &#039;&#039;&#039;Stadium → preamp input → preamp output → Stadium&#039;&#039;&#039;<br /> <br /> For an amplifier preamp:<br /> <br /> &#039;&#039;&#039;Stadium → amplifier input → PREAMP OUT or FX SEND → Stadium&#039;&#039;&#039;<br /> <br /> The amplifier must provide an appropriate preamp-level output such as:<br /> <br /> * PREAMP OUT<br /> * FX SEND<br /> <br /> This allows the preamp stage to be measured independently from the power amplifier and cabinet.<br /> <br /> == Distortion Clone ==<br /> <br /> A &#039;&#039;&#039;Distortion&#039;&#039;&#039; Clone captures a gain-based pedal.<br /> <br /> Supported targets include devices such as:<br /> <br /> * Distortion<br /> * Overdrive<br /> * Boost<br /> * Fuzz<br /> <br /> A typical path is:<br /> <br /> &#039;&#039;&#039;Stadium → pedal → Stadium&#039;&#039;&#039;<br /> <br /> The resulting Clone can then be inserted into a Stadium preset as a Distortion-type Clone block.<br /> <br /> Despite the category name &quot;Distortion,&quot; the target does not have to be strictly a conventional distortion pedal.<br /> <br /> Line 6 explicitly includes overdrive, boost, and fuzz devices within this category.<br /> <br /> == Combined gain chains ==<br /> <br /> Proxy can also capture more than one gain-producing device as part of an Amp+Cab signal chain.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Stadium → drive pedal → amplifier → cabinet → microphone → Stadium&#039;&#039;&#039;<br /> <br /> Line 6 states that one or more drive pedals may be placed before the amplifier when creating an Amp+Cab Clone.<br /> <br /> This makes it possible to clone an entire gain-based signal flow rather than only an isolated amplifier.<br /> <br /> The resulting Clone represents the combined behavior of those devices.<br /> <br /> The individual pedal and amplifier cannot then be independently separated from that Clone.<br /> <br /> See [[Capture Types]].<br /> <br /> == Devices that should not be included ==<br /> <br /> Line 6 recommends avoiding compression, modulation, and time-based effects in the signal path during Proxy cloning.<br /> <br /> Examples include:<br /> <br /> * Compressor<br /> * Chorus<br /> * Flanger<br /> * Delay<br /> * Reverb<br /> * Tremolo<br /> * Other time-varying effects<br /> <br /> Amplifier reverb, tremolo, and similar onboard effects should normally be disabled.<br /> <br /> The goal is to capture the core nonlinear tone of the target equipment rather than effects whose behavior may not be represented appropriately by the Proxy cloning process.<br /> <br /> == Cloud-based training ==<br /> <br /> Proxy model creation is performed using Line 6 online servers.<br /> <br /> The physical measurement occurs through Helix Stadium, but the collected data is uploaded for processing.<br /> <br /> This means creating a Proxy Clone requires:<br /> <br /> * Helix Stadium<br /> * Internet connection<br /> * Connection to Wi-Fi or Ethernet<br /> * Line6.com account<br /> * Sign-in to that account<br /> <br /> The cloud servers perform the computationally intensive model creation.<br /> <br /> The completed Clone is then returned to the Stadium.<br /> <br /> == Why cloud processing is used ==<br /> <br /> Line 6 states that cloud processing allows Proxy to use substantially greater computing resources than would normally be available inside the floor processor.<br /> <br /> This can allow more detailed training without requiring the Stadium hardware itself to perform the complete machine-learning process.<br /> <br /> It also separates the model-training technology from the firmware running on the floor unit.<br /> <br /> Line 6 states that this architecture allows the underlying Proxy technology to be improved on the server side without every improvement necessarily requiring a corresponding firmware update.<br /> <br /> == Creating a Clone ==<br /> <br /> The current Stadium workflow begins from:<br /> <br /> &#039;&#039;&#039;Main Menu → New Clone&#039;&#039;&#039;<br /> <br /> The user then selects the target type:<br /> <br /> * Amp+Cab<br /> * Amp<br /> * Preamp<br /> * Distortion<br /> <br /> Stadium provides on-screen connection and setup instructions appropriate for the selected target.<br /> <br /> The exact SEND and RETURN connections differ between Helix Stadium and Helix Stadium XL, so the on-screen instructions and current manual should be followed rather than assuming one universal jack configuration.<br /> <br /> == Level setup ==<br /> <br /> The Proxy workflow guides the user through level configuration before cloning begins.<br /> <br /> Correct levels are important for two reasons:<br /> <br /> * Avoiding unwanted clipping or noise<br /> * Presenting the target device with an appropriate signal level<br /> <br /> For nonlinear equipment, the level entering the device affects its behavior.<br /> <br /> A distorted amplifier or pedal may therefore respond differently if the Proxy test signal drives it substantially harder or softer than intended.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Noise and grounding ==<br /> <br /> Line 6 recommends making the cloning signal path as noise-free as practical.<br /> <br /> Potential problems include:<br /> <br /> * Ground loops<br /> * Computer connections<br /> * USB connections<br /> * MIDI connections<br /> * Power supplies<br /> * Unbalanced audio connections<br /> <br /> Line 6 specifically recommends disconnecting unnecessary MIDI and USB-C cables during the cloning process to reduce possible interference.<br /> <br /> Appropriate audio isolation or ground-lift facilities on external equipment may also help resolve ground-loop noise.<br /> <br /> &#039;&#039;&#039;Never defeat protective mains safety grounding to eliminate hum.&#039;&#039;&#039;<br /> <br /> == Collecting Device DNA ==<br /> <br /> During cloning, Stadium sends test signals through the target equipment.<br /> <br /> Line 6 describes this stage as:<br /> <br /> &#039;&#039;&#039;Collecting Device DNA&#039;&#039;&#039;<br /> <br /> The user should leave the target equipment controls unchanged while the measurement process is running.<br /> <br /> Changing:<br /> <br /> * Gain<br /> * EQ<br /> * Channel<br /> * Pedal controls<br /> * Amplifier controls<br /> <br /> during measurement can invalidate the relationship Proxy is attempting to learn.<br /> <br /> == Cloud processing ==<br /> <br /> After the physical measurement is completed, the captured data is uploaded to Line 6 servers.<br /> <br /> The server analyzes the measurement and creates the Proxy Clone.<br /> <br /> This separation means the Stadium itself does not have to perform the entire training process locally.<br /> <br /> Once processing completes, the resulting Clone is returned for comparison.<br /> <br /> == Comparing the Clone ==<br /> <br /> Proxy provides a comparison stage after model creation.<br /> <br /> The user can compare:<br /> <br /> * The original physical equipment<br /> * The resulting Clone<br /> <br /> For appropriate Clone types, Stadium can also provide Amp or Cab blocks during the comparison process so the Clone can be heard in a useful playback context.<br /> <br /> These comparison blocks are for auditioning and are not automatically stored as part of the Clone file.<br /> <br /> Listen for differences in:<br /> <br /> * Gain<br /> * Frequency response<br /> * Pick attack<br /> * Dynamics<br /> * Compression<br /> * Low-frequency response<br /> * High-frequency response<br /> * Sustain<br /> * Cleanup<br /> <br /> Level matching is important when comparing the Clone and original equipment.<br /> <br /> == Saving a Clone ==<br /> <br /> If the result is satisfactory, the Clone can be saved to the Stadium Clone Library.<br /> <br /> The creator enters a Clone name and can provide metadata including:<br /> <br /> * Brand<br /> * Product<br /> * Notes<br /> <br /> Notes can document information such as:<br /> <br /> * Amplifier channel<br /> * Gain<br /> * EQ settings<br /> * Pedal settings<br /> * Intended use<br /> * Other capture information<br /> <br /> This metadata is also used by Stadium&#039;s search system.<br /> <br /> Good metadata becomes increasingly important as the user&#039;s Clone library grows.<br /> <br /> == Clone Library ==<br /> <br /> Helix Stadium can store up to &#039;&#039;&#039;2,048 Clones&#039;&#039;&#039; internally.<br /> <br /> Clones can be:<br /> <br /> * Created<br /> * Imported<br /> * Exported<br /> * Browsed<br /> * Searched<br /> * Added to presets<br /> <br /> Clone files are managed similarly to other user content within the Stadium ecosystem.<br /> <br /> == Clone blocks ==<br /> <br /> Clones appear in the appropriate Stadium model categories.<br /> <br /> An:<br /> <br /> &#039;&#039;&#039;Amp Clone&#039;&#039;&#039;<br /> <br /> appears in the Amp category.<br /> <br /> An:<br /> <br /> &#039;&#039;&#039;Amp+Cab Clone&#039;&#039;&#039;<br /> <br /> also appears in the Amp category.<br /> <br /> A:<br /> <br /> &#039;&#039;&#039;Preamp Clone&#039;&#039;&#039;<br /> <br /> appears in the Preamp category.<br /> <br /> A:<br /> <br /> &#039;&#039;&#039;Distortion Clone&#039;&#039;&#039;<br /> <br /> appears in the Distortion category.<br /> <br /> This allows Clones to be used alongside Line 6&#039;s conventional modeled blocks.<br /> <br /> == Clone controls ==<br /> <br /> Clone blocks provide controls similar to other Stadium Amp and Effect models.<br /> <br /> Available controls include Gain and EQ parameters.<br /> <br /> Clone parameters can also participate in Stadium features such as:<br /> <br /> * Snapshots<br /> * Stomp control<br /> * External pedals<br /> * External switches<br /> * Triggers<br /> * MIDI control<br /> * User Defaults<br /> * Favorites<br /> <br /> The existence of these controls should not automatically be interpreted as meaning that Proxy recreated every physical control circuit of the original device.<br /> <br /> The Clone fundamentally represents the measured device configuration.<br /> <br /> == DSP usage ==<br /> <br /> Proxy Clones require significant DSP resources.<br /> <br /> Line 6 states that a Clone block can require approximately the same DSP resources as an &#039;&#039;&#039;Agoura&#039;&#039;&#039; Amp model.<br /> <br /> A Stadium preset can contain up to:<br /> <br /> &#039;&#039;&#039;four Clone blocks&#039;&#039;&#039;<br /> <br /> with a maximum of:<br /> <br /> &#039;&#039;&#039;two Clones per main path&#039;&#039;&#039;<br /> <br /> subject to the overall DSP requirements of the preset.<br /> <br /> == Mono processing ==<br /> <br /> All Proxy Clone blocks are currently &#039;&#039;&#039;mono&#039;&#039;&#039;.<br /> <br /> If a stereo block is placed before a Clone, the signal entering the Clone is summed to mono.<br /> <br /> This is important when designing Stadium presets containing stereo effects.<br /> <br /> Stereo processing intended to remain stereo should generally be placed after the Clone unless there is a specific reason to sum it beforehand.<br /> <br /> == Cabinets with Amp Clones ==<br /> <br /> An Amp Clone does not include a speaker cabinet.<br /> <br /> When used through a full-range system, a Cab block or IR will normally be placed after it:<br /> <br /> &#039;&#039;&#039;Amp Clone → Cab/IR → output&#039;&#039;&#039;<br /> <br /> Line 6&#039;s Model List workflow specifically notes that an Amp Clone will typically be followed manually by a Cab block.<br /> <br /> An Amp+Cab Clone already contains the measured physical cabinet and microphone.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Physical guitar cabinet playback ==<br /> <br /> A direct Amp Clone can also be used with an appropriate physical guitar cabinet.<br /> <br /> A possible signal chain is:<br /> <br /> &#039;&#039;&#039;Instrument → Amp Clone → suitable power amplifier → physical guitar cabinet&#039;&#039;&#039;<br /> <br /> In this situation, additional virtual cabinet processing is normally unnecessary because the physical guitar cabinet supplies the speaker and cabinet response.<br /> <br /> == CustomTone ==<br /> <br /> Proxy Clones can be shared through Line 6&#039;s &#039;&#039;&#039;CustomTone&#039;&#039;&#039; community.<br /> <br /> Users can upload and download:<br /> <br /> * Helix Stadium presets<br /> * Proxy Clones<br /> <br /> This allows Clones created by one Stadium owner to be used by other compatible systems.<br /> <br /> As with other community capture libraries, useful metadata makes shared models much easier to evaluate and organize.<br /> <br /> == Import and export ==<br /> <br /> The Helix Stadium application provides librarian functions for managing Clones.<br /> <br /> Users can:<br /> <br /> * Import Clone files<br /> * Export Clone files<br /> * Browse the Clone Library<br /> * Organize content<br /> <br /> This allows Clone collections to be backed up and transferred independently of individual presets.<br /> <br /> == Helix Stadium Native ==<br /> <br /> &#039;&#039;&#039;Helix Stadium Native&#039;&#039;&#039; supports Proxy Clone loading.<br /> <br /> This allows Proxy Clones to be used in a DAW environment as well as on Stadium hardware.<br /> <br /> The plugin includes the Stadium amp, cabinet, and effects environment together with:<br /> <br /> * Cabinet IR support<br /> * Proxy Clone loading<br /> * Preset compatibility with Stadium hardware<br /> <br /> This extends Proxy playback beyond the physical floor processor.<br /> <br /> == Creating versus playing Clones ==<br /> <br /> There is an important distinction between &#039;&#039;&#039;creating&#039;&#039;&#039; and &#039;&#039;&#039;playing&#039;&#039;&#039; Proxy Clones.<br /> <br /> Creating a Clone currently requires compatible Helix Stadium hardware because the Stadium performs the physical measurement and interfaces with the cloud cloning system.<br /> <br /> Once created, Clones can also be loaded by compatible Line 6 software such as Helix Stadium Native.<br /> <br /> A system capable of playing a Clone should therefore not automatically be assumed capable of creating one.<br /> <br /> == Proxy and Agoura ==<br /> <br /> Proxy should not be confused with Line 6&#039;s &#039;&#039;&#039;Agoura&#039;&#039;&#039; amplifier modeling.<br /> <br /> Agoura is Line 6&#039;s conventional amplifier-modeling methodology used to create detailed predefined amplifier models for Helix Stadium.<br /> <br /> Proxy is a user-driven capture technology.<br /> <br /> The two can coexist in the same Stadium preset:<br /> <br /> &#039;&#039;&#039;Agoura model = Line 6-developed amplifier model&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Proxy Clone = measured model created from physical user equipment&#039;&#039;&#039;<br /> <br /> This distinction is similar to the difference between algorithmic amplifier models and user-created captures on other platforms.<br /> <br /> == Proxy and legacy Helix modeling ==<br /> <br /> Helix Stadium also retains compatibility with the established Helix/HX amplifier and effects ecosystem.<br /> <br /> Proxy therefore does not replace conventional Line 6 modeling.<br /> <br /> A Stadium preset can combine:<br /> <br /> * Legacy Helix/HX models<br /> * Agoura models<br /> * Proxy Clones<br /> * Cabinet models<br /> * IRs<br /> * Effects<br /> <br /> This makes capture technology one component of a broader modeling and effects platform.<br /> <br /> == Proxy and NAM ==<br /> <br /> Proxy and [[Neural Amp Modeler]] both use measured data to reproduce physical audio equipment, but they are separate technologies.<br /> <br /> A Proxy Clone is not a &#039;&#039;.nam&#039;&#039; file.<br /> <br /> The systems differ in:<br /> <br /> * Capture signal<br /> * Training process<br /> * Model architecture<br /> * Model format<br /> * Playback engine<br /> * Distribution<br /> * Hardware integration<br /> <br /> NAM provides an open-source model and playback ecosystem.<br /> <br /> Proxy is a proprietary Line 6 technology integrated with Helix Stadium and Line 6 cloud infrastructure.<br /> <br /> == Proxy and Neural Capture V2 ==<br /> <br /> Proxy and [[Neural Capture]] V2 share one broad architectural concept:<br /> <br /> &#039;&#039;&#039;physical measurement on hardware + computationally intensive cloud training&#039;&#039;&#039;<br /> <br /> The implementations remain independent.<br /> <br /> Neural Capture V2 uses Cortex hardware, Cortex Control, and Cortex Cloud.<br /> <br /> Proxy uses Helix Stadium hardware and Line 6 online servers.<br /> <br /> Neither model format is interchangeable with the other.<br /> <br /> == Proxy and TONEX ==<br /> <br /> Proxy and [[TONEX]] can both create models of amplifiers, pedals, and complete amplifier/cabinet rigs.<br /> <br /> The model formats and playback systems are different.<br /> <br /> TONEX Modeler can preserve physical capture data and perform training separately, while Proxy currently integrates the physical measurement and cloud-training workflow directly through Helix Stadium.<br /> <br /> == Phase 1 ==<br /> <br /> Line 6 describes the Proxy implementation introduced in Stadium 1.3 as &#039;&#039;&#039;Phase 1&#039;&#039;&#039;.<br /> <br /> This wording is significant.<br /> <br /> The initial release supports:<br /> <br /> * Amp+Cab<br /> * Amp<br /> * Preamp<br /> * Distortion<br /> <br /> but should not necessarily be treated as the final scope or capability of Proxy.<br /> <br /> Future firmware, cloud technology, target types, or workflow capabilities may expand the system.<br /> <br /> Documentation should therefore distinguish between the current Proxy implementation and assumptions about future development.<br /> <br /> == Troubleshooting ==<br /> <br /> If a Proxy Clone does not reproduce the expected sound, check:<br /> <br /> * Device connections<br /> * Clone type<br /> * Send and return levels<br /> * Load-box configuration<br /> * Microphone path<br /> * Grounding<br /> * Noise<br /> * Unwanted effects<br /> * Amplifier settings<br /> * Cabinet configuration<br /> * Comparison level<br /> <br /> If cloning cannot begin or cloud processing fails, also check:<br /> <br /> * Internet connection<br /> * Wi-Fi or Ethernet<br /> * Line6.com account login<br /> * Stadium firmware<br /> * Line 6 service availability<br /> <br /> If a Clone sounds correct during comparison but different in a preset, check:<br /> <br /> * Cab block<br /> * IR<br /> * Input level<br /> * Additional effects<br /> * EQ<br /> * Signal routing<br /> * Stereo-to-mono summing<br /> * Output configuration<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Official resources ==<br /> <br /> * [https://manuals.line6.com/en/helix-stadium/live/clones Line 6 Proxy Cloning documentation]<br /> * [https://line6.com/helix-stadium/ Helix Stadium]<br /> * [https://line6.com/customtone/ Line 6 CustomTone]<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Neural Amp Modeler]]<br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Neural Capture]]<br /> * [[Capture Types]]<br /> * [[Creating a Capture]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss Proxy Clones, Helix Stadium, and cloning workflows on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Line 6 Proxy]]<br /> [[Category:Capture technology]]<br /> [[Category:Commercial capture systems]]</div> Tue, 01 Sep 2026 06:06:36 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Line_6_Proxy https://namforum.com/wiki/index.php?title=Neural_Capture&diff=18&oldid=0 https://namforum.com/wiki/index.php?title=Neural_Capture&diff=18&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Neural Capture =<br /> <br /> &#039;&#039;&#039;Neural Capture&#039;&#039;&#039; is Neural DSP&#039;s technology for creating digital models of physical amplifiers, cabinets, pedals, and other compatible audio equipment.<br /> <br /> Neural Captures are used within Neural DSP&#039;s Cortex hardware ecosystem and can be shared through Cortex Cloud.<br /> <br /> There are currently two generations of the technology:<br /> <br /> * &#039;&#039;&#039;Neural Capture Version 1&#039;&#039;&#039;<br /> * &#039;&#039;&#039;Neural Capture Version 2&#039;&#039;&#039;<br /> <br /> Both create playable Capture blocks, but they differ substantially in training method, computational requirements, and the types of equipment for which they are optimized.<br /> <br /> == Overview ==<br /> <br /> A Neural Capture measures the response of physical equipment to test signals generated by the Cortex system.<br /> <br /> Depending on the target equipment, a Capture can represent:<br /> <br /> * Amplifier<br /> * Amplifier and cabinet<br /> * Combo amplifier<br /> * Cabinet<br /> * Overdrive or distortion device<br /> * Fuzz<br /> * Compressor<br /> * Larger signal chain<br /> <br /> Once created, a Neural Capture can be inserted as a block in a Cortex signal chain.<br /> <br /> == Basic capture process ==<br /> <br /> A simplified capture path is:<br /> <br /> &#039;&#039;&#039;Cortex capture signal → device under test → Cortex return&#039;&#039;&#039;<br /> <br /> For a pedal:<br /> <br /> &#039;&#039;&#039;Cortex → pedal → Cortex&#039;&#039;&#039;<br /> <br /> For a direct amplifier:<br /> <br /> &#039;&#039;&#039;Cortex → amplifier → suitable load/direct-output device → Cortex&#039;&#039;&#039;<br /> <br /> For a miked amplifier and cabinet:<br /> <br /> &#039;&#039;&#039;Cortex → amplifier → cabinet → microphone → Cortex&#039;&#039;&#039;<br /> <br /> The Cortex system sends test signals through the equipment and records its response.<br /> <br /> The captured data is then used to create the Neural Capture model.<br /> <br /> == Neural Capture Version 1 ==<br /> <br /> &#039;&#039;&#039;Neural Capture Version 1&#039;&#039;&#039; is the original Neural DSP capture algorithm.<br /> <br /> V1 training occurs directly on compatible Cortex hardware.<br /> <br /> It is:<br /> <br /> * Fast<br /> * Computationally efficient<br /> * Fully offline<br /> * Suitable for most amplifiers<br /> * Suitable for cabinets<br /> * Suitable for many overdrive pedals<br /> <br /> A typical V1 Capture takes only a few minutes to create.<br /> <br /> Because training happens locally, no cloud connection is required for the modeling process.<br /> <br /> V1 remains fully supported alongside Version 2.<br /> <br /> == Neural Capture Version 2 ==<br /> <br /> &#039;&#039;&#039;Neural Capture Version 2&#039;&#039;&#039; was introduced in November 2025.<br /> <br /> V2 uses a newer, higher-resolution capture algorithm designed to reproduce more complex dynamic behavior.<br /> <br /> Unlike V1, the training process does not occur entirely on the Cortex hardware.<br /> <br /> Instead:<br /> <br /> # Cortex hardware performs the physical measurement.<br /> # Cortex Control manages the capture session.<br /> # Captured data is sent to Cortex Cloud.<br /> # Cortex Cloud performs the higher-resolution training.<br /> # The completed Neural Capture is returned to the Cortex ecosystem.<br /> <br /> An internet connection and Neural DSP account are therefore required when creating V2 Captures.<br /> <br /> == Why V2 was developed ==<br /> <br /> Some analog devices are particularly difficult to reproduce using capture systems.<br /> <br /> Their behavior may depend strongly on:<br /> <br /> * Input amplitude<br /> * Playing dynamics<br /> * Transients<br /> * Previous signal history<br /> * Compression<br /> * Power-supply behavior<br /> * Interaction between controls<br /> <br /> Neural DSP developed Capture V2 specifically to improve reproduction of these characteristics.<br /> <br /> Examples include:<br /> <br /> * Vintage fuzz pedals<br /> * Compressors<br /> * Highly dynamic tube amplifiers<br /> * Amplifiers exhibiting sag and bloom<br /> * Devices with blend controls<br /> <br /> V2 is not intended merely as a replacement with a higher version number.<br /> <br /> It represents a more computationally demanding option for cases where additional dynamic accuracy is useful.<br /> <br /> == V1 versus V2 ==<br /> <br /> Both capture generations remain useful.<br /> <br /> &#039;&#039;&#039;Version 1&#039;&#039;&#039; is generally appropriate when:<br /> <br /> * Offline capture is required<br /> * Fast capture is important<br /> * DSP resources are limited<br /> * The target is a conventional amplifier<br /> * The target is a cabinet<br /> * The target is a relatively straightforward overdrive<br /> <br /> &#039;&#039;&#039;Version 2&#039;&#039;&#039; is particularly useful when:<br /> <br /> * Maximum dynamic detail is desired<br /> * Capturing fuzz<br /> * Capturing compression<br /> * Capturing a highly touch-sensitive amplifier<br /> * Capturing power-amp sag or bloom<br /> * Capturing complex dynamic behavior<br /> <br /> V2 models use more processing resources during playback than V1 models.<br /> <br /> The choice therefore involves both modeling requirements and available DSP.<br /> <br /> == V2 capture time ==<br /> <br /> Neural DSP states that a V2 Capture session typically takes approximately &#039;&#039;&#039;10 minutes&#039;&#039;&#039;.<br /> <br /> This includes the cloud-based training process.<br /> <br /> Actual completion time can depend on the capture session and network/cloud conditions.<br /> <br /> V1 is substantially faster because its lighter training algorithm runs locally.<br /> <br /> == Creating V2 Captures ==<br /> <br /> The current V2 workflow uses &#039;&#039;&#039;Cortex Control&#039;&#039;&#039;.<br /> <br /> A typical procedure is:<br /> <br /> # Connect the target equipment according to the displayed connection diagram.<br /> # Open Cortex Control.<br /> # Log into the Neural DSP account.<br /> # Select &#039;&#039;&#039;New Neural Capture&#039;&#039;&#039;.<br /> # Select &#039;&#039;&#039;Neural Capture Version 2&#039;&#039;&#039;.<br /> # Configure the capture connections.<br /> # Calibrate levels.<br /> # Enter the required metadata.<br /> # Start the capture.<br /> # Wait for cloud processing.<br /> # A/B the resulting Capture against the original equipment.<br /> # Save the Capture.<br /> <br /> Cortex Control acts as the connection between the Cortex device and Cortex Cloud during V2 creation.<br /> <br /> == V2 metadata ==<br /> <br /> Metadata has an unusually important role in Neural Capture V2.<br /> <br /> Before V2 training begins, the creator identifies the type of equipment being captured.<br /> <br /> Current V2 device categories include:<br /> <br /> * Default<br /> * Amp<br /> * Amp + Cab<br /> * Combo Amp<br /> * Cab<br /> * Overdrive<br /> * Fuzz<br /> * Compressor<br /> <br /> Neural DSP states that Cortex Cloud uses this information during processing.<br /> <br /> Selecting the correct target type is therefore part of the modeling process rather than merely library organization.<br /> <br /> == Capture calibration ==<br /> <br /> The Neural Capture workflow includes a calibration stage.<br /> <br /> Correct send and return levels are important because the capture system must measure the device over an appropriate operating range.<br /> <br /> For nonlinear equipment, changing the signal level entering the device can change:<br /> <br /> * Distortion<br /> * Compression<br /> * Dynamic response<br /> * Attack<br /> * Sustain<br /> * Cleanup behavior<br /> <br /> Do not treat calibration as merely a way to prevent clipping.<br /> <br /> It determines the conditions under which the device is measured.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == A/B testing ==<br /> <br /> After the capture is completed, the Cortex workflow provides an A/B comparison between the original equipment and the resulting Neural Capture.<br /> <br /> This allows the creator to evaluate the model before saving it.<br /> <br /> Listen for differences in:<br /> <br /> * Gain<br /> * Frequency response<br /> * Pick attack<br /> * Dynamics<br /> * Compression<br /> * Sustain<br /> * Cleanup<br /> * Low-frequency behavior<br /> * High-frequency behavior<br /> <br /> Comparisons should be made at closely matched levels.<br /> <br /> A louder signal can easily be perceived as fuller or better even when the underlying model is not more accurate.<br /> <br /> == Capturing amplifiers ==<br /> <br /> An amplifier can be captured directly without its physical cabinet response.<br /> <br /> This requires a safe and appropriate method of obtaining the amplifier output.<br /> <br /> A typical path is:<br /> <br /> &#039;&#039;&#039;Cortex → amplifier → reactive load/direct-output device → Cortex&#039;&#039;&#039;<br /> <br /> The resulting Capture can then be combined with:<br /> <br /> * Cortex cabinet processing<br /> * An impulse response<br /> * A physical guitar cabinet<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Tube amplifier safety ==<br /> <br /> &#039;&#039;&#039;Never connect the speaker output of a tube amplifier directly to an ordinary Cortex input.&#039;&#039;&#039;<br /> <br /> Doing so can damage both the amplifier and the Cortex device.<br /> <br /> When capturing a tube amplifier, use:<br /> <br /> * A suitable D.I. output while the amplifier remains connected to the required speaker load, or<br /> * An appropriate reactive load box between the amplifier speaker output and the Cortex input<br /> <br /> The amplifier must always see the load required for safe operation.<br /> <br /> == Capturing amplifier and cabinet ==<br /> <br /> A complete amplifier/cabinet signal chain can be captured using a microphone.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Cortex → amplifier → cabinet → microphone → Cortex&#039;&#039;&#039;<br /> <br /> The resulting Neural Capture can then include the combined behavior of:<br /> <br /> * Amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Return path<br /> <br /> Additional cabinet processing may not be required when the cabinet and microphone are already part of the Capture.<br /> <br /> See [[Capture Types]].<br /> <br /> == Cabinet captures ==<br /> <br /> Neural Capture can also capture a cabinet independently.<br /> <br /> A cabinet Capture represents the measured speaker/cabinet/microphone path used during creation.<br /> <br /> This provides an alternative to conventional static cabinet impulse responses within the Cortex environment.<br /> <br /> Cabinet Captures and conventional IRs should not automatically be assumed to be technically identical simply because they can serve a similar role in a guitar signal chain.<br /> <br /> == Pedal captures ==<br /> <br /> Neural Capture can reproduce compatible nonlinear pedals.<br /> <br /> V1 has traditionally been used for devices such as:<br /> <br /> * Overdrive<br /> * Distortion<br /> * Boost<br /> <br /> V2 extends the practical range toward more dynamically difficult devices such as:<br /> <br /> * Fuzz<br /> * Compressor<br /> <br /> The physical pedal settings used during capture form part of the operating point represented by the resulting model.<br /> <br /> For substantially different settings, separate Captures may be useful.<br /> <br /> == Fuzz capture ==<br /> <br /> Fuzz circuits can present a particularly difficult modeling problem.<br /> <br /> Some fuzz pedals interact strongly with:<br /> <br /> * Source impedance<br /> * Guitar volume control<br /> * Pickup<br /> * Input level<br /> * Circuit bias<br /> * Playing dynamics<br /> <br /> Neural DSP identifies fuzz as one of the primary applications for Capture V2.<br /> <br /> V2 is specifically intended to improve behaviors such as guitar-volume cleanup that can be difficult for simpler capture systems to reproduce.<br /> <br /> == Compressor capture ==<br /> <br /> Capture V2 adds explicit support for compressors.<br /> <br /> Compression is inherently dynamic because the processor&#039;s behavior changes according to the level and timing of the incoming signal.<br /> <br /> Neural DSP notes that V2 can capture compressors with high accuracy, although extreme threshold settings and long release times may produce greater differences from the original hardware.<br /> <br /> This illustrates an important principle:<br /> <br /> No capture architecture should automatically be assumed to reproduce every possible device and setting perfectly.<br /> <br /> The result should always be evaluated against the original.<br /> <br /> == Neural Capture blocks ==<br /> <br /> Once created or downloaded, a Neural Capture can be placed on &#039;&#039;&#039;The Grid&#039;&#039;&#039; as a processing block.<br /> <br /> V1 and V2 Captures are visually distinguishable.<br /> <br /> V2 Capture blocks use a different double-layered outline.<br /> <br /> Both versions provide the same general parameter-editor layout.<br /> <br /> == Cortex Cloud ==<br /> <br /> &#039;&#039;&#039;Cortex Cloud&#039;&#039;&#039; is Neural DSP&#039;s online system for storing, sharing, and discovering Cortex content.<br /> <br /> Users can make Captures available to other Cortex users.<br /> <br /> V2 relies more directly on Cortex Cloud because its training process occurs there.<br /> <br /> V2 Captures are automatically stored privately in the creator&#039;s Cortex Cloud profile during creation.<br /> <br /> They can subsequently be managed and shared according to the available Cortex Cloud controls.<br /> <br /> == Sharing Captures ==<br /> <br /> Community Neural Captures can be downloaded from Cortex Cloud and used in compatible Cortex hardware.<br /> <br /> Useful capture information includes:<br /> <br /> * Creator<br /> * Device manufacturer<br /> * Device model<br /> * Capture type<br /> * Device settings<br /> * Cabinet<br /> * Microphone<br /> * Gain<br /> * Instrument<br /> * Notes<br /> <br /> Good metadata is particularly valuable when a Capture becomes separated from the discussion or session in which it was originally created.<br /> <br /> == Quad Cortex ==<br /> <br /> &#039;&#039;&#039;Quad Cortex&#039;&#039;&#039; is Neural DSP&#039;s flagship Cortex hardware platform.<br /> <br /> It combines:<br /> <br /> * Neural Capture playback<br /> * Neural Capture creation<br /> * Algorithmic amplifier models<br /> * Cabinet modeling<br /> * Effects<br /> * Routing<br /> * MIDI<br /> * USB audio<br /> * Preset management<br /> <br /> Quad Cortex can create and play both Neural Capture V1 and V2.<br /> <br /> V1 can be created directly on the hardware.<br /> <br /> V2 creation uses Quad Cortex together with Cortex Control and Cortex Cloud.<br /> <br /> == Cortex Control ==<br /> <br /> &#039;&#039;&#039;Cortex Control&#039;&#039;&#039; is Neural DSP&#039;s desktop application for managing Cortex hardware.<br /> <br /> It provides computer-based control over functions including:<br /> <br /> * Preset editing<br /> * Device management<br /> * Cortex Cloud interaction<br /> * Neural Capture creation<br /> <br /> The complete V2 capture process can be initiated and managed through Cortex Control.<br /> <br /> V2 creation requires Cortex Control because the application connects the hardware capture process with cloud training.<br /> <br /> == Nano Cortex ==<br /> <br /> &#039;&#039;&#039;Nano Cortex&#039;&#039;&#039; is Neural DSP&#039;s compact Cortex hardware platform.<br /> <br /> Nano Cortex supports Neural Capture playback and V1 capture functionality.<br /> <br /> With NanOS 2.2.0, Neural DSP added the &#039;&#039;&#039;Capture 2 Player&#039;&#039;&#039;, allowing Nano Cortex to load and play V2 Captures created using Quad Cortex.<br /> <br /> At the introduction of Capture V2, direct creation of V2 Captures on Nano Cortex was not yet available.<br /> <br /> Because this capability is under active development, consult current Neural DSP documentation for its latest status.<br /> <br /> == Quad Cortex mini ==<br /> <br /> The &#039;&#039;&#039;Quad Cortex mini&#039;&#039;&#039; extends Neural DSP&#039;s Cortex hardware family with a smaller form factor while retaining the broader Cortex ecosystem.<br /> <br /> Its exact capture and playback capabilities should be checked against current Neural DSP documentation because Cortex hardware and software features continue to evolve.<br /> <br /> == Factory Captures ==<br /> <br /> Neural DSP supplies factory Neural Captures in addition to community-created content.<br /> <br /> When Capture V2 was introduced, Neural DSP added &#039;&#039;&#039;669 V2 Captures across 41 devices&#039;&#039;&#039; to its library and official Cortex Cloud profile.<br /> <br /> These included examples of:<br /> <br /> * Guitar amplifiers<br /> * Combo amplifiers<br /> * Overdrives<br /> * Fuzz pedals<br /> * Compressors<br /> <br /> Factory libraries continue to evolve, so exact counts should be treated as historical information rather than permanent specifications.<br /> <br /> == Neural Capture and conventional Neural DSP models ==<br /> <br /> Neural Capture should not be confused with Neural DSP&#039;s conventional algorithmic device models.<br /> <br /> Quad Cortex contains both.<br /> <br /> An algorithmic amplifier model is developed by Neural DSP to represent an amplifier and normally exposes modeled controls.<br /> <br /> A Neural Capture is created by measuring a particular physical device or rig at a particular configuration.<br /> <br /> They are different technologies that can coexist in the same Cortex preset.<br /> <br /> == Neural Capture and NAM ==<br /> <br /> Neural Capture and [[Neural Amp Modeler]] are both data-driven approaches to reproducing physical audio equipment, but they are separate technologies.<br /> <br /> A Neural Capture is not a &#039;&#039;.nam&#039;&#039; file.<br /> <br /> The ecosystems differ in:<br /> <br /> * Capture process<br /> * Training<br /> * Model representation<br /> * Playback engine<br /> * Distribution<br /> * Hardware integration<br /> <br /> NAM provides an open-source model and playback ecosystem.<br /> <br /> Neural Capture is Neural DSP&#039;s proprietary technology integrated with Cortex hardware and Cortex Cloud.<br /> <br /> == Neural Capture and TONEX ==<br /> <br /> Neural Capture and [[TONEX]] can both create models of amplifiers, pedals, cabinets, and complete rigs.<br /> <br /> However, their model formats and playback systems are not interchangeable.<br /> <br /> Both systems also now distinguish the physical capture process from computationally intensive training in some workflows:<br /> <br /> * TONEX Modeler can preserve capture data and train it later.<br /> * Neural Capture V2 sends captured data to Cortex Cloud for training.<br /> <br /> The engineering implementations remain different.<br /> <br /> == V1 and V2 compatibility ==<br /> <br /> V1 remains supported after the introduction of V2.<br /> <br /> This is important because V2 is more computationally expensive and requires cloud processing during creation.<br /> <br /> The two systems therefore serve different practical purposes rather than V2 simply making V1 obsolete.<br /> <br /> A user may deliberately choose V1 for:<br /> <br /> * Faster capture<br /> * Offline operation<br /> * Lower DSP usage<br /> <br /> and V2 for:<br /> <br /> * Greater dynamic detail<br /> * Fuzz<br /> * Compression<br /> * Highly responsive amplifiers<br /> <br /> == Troubleshooting ==<br /> <br /> If a Neural Capture differs substantially from the original equipment, check:<br /> <br /> * Connections<br /> * Send level<br /> * Return level<br /> * Calibration<br /> * Clipping<br /> * Load-box configuration<br /> * Microphone path<br /> * Capture type<br /> * Metadata selection for V2<br /> * Additional cabinet processing<br /> * A/B comparison level<br /> <br /> For V2, also check:<br /> <br /> * Internet connection<br /> * Cortex Control<br /> * Neural DSP account login<br /> * Audio-driver configuration<br /> * Software permissions<br /> * Other applications using the audio device<br /> <br /> A DAW or other program accessing the same audio system can interfere with a V2 capture session.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Official resources ==<br /> <br /> * [https://neuraldsp.com/manual/quad-cortex Quad Cortex User Manual]<br /> * [https://neuraldsp.com/news/introducing-neural-capture-version-2 Introducing Neural Capture Version 2]<br /> * [https://neuraldsp.com/quad-cortex Quad Cortex]<br /> * [https://neuraldsp.com/cortex-control Cortex Control]<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Neural Amp Modeler]]<br /> * [[TONEX]]<br /> * [[Kemper Profiling]]<br /> * [[Capture Types]]<br /> * [[Creating a Capture]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss Neural Capture, Quad Cortex, and Cortex workflows on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Neural Capture]]<br /> [[Category:Capture technology]]<br /> [[Category:Commercial capture systems]]</div> Tue, 01 Sep 2026 06:05:49 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Neural_Capture https://namforum.com/wiki/index.php?title=Kemper_Profiling&diff=17&oldid=0 https://namforum.com/wiki/index.php?title=Kemper_Profiling&diff=17&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Kemper Profiling =<br /> <br /> &#039;&#039;&#039;Kemper Profiling&#039;&#039;&#039; is the capture technology used by the &#039;&#039;&#039;KEMPER PROFILER&#039;&#039;&#039; family of guitar and bass processors.<br /> <br /> Introduced by Kemper as an alternative to conventional predefined amplifier modeling, PROFILING allows a user to measure a real amplifier or amplifier-based signal chain and create a digital &#039;&#039;&#039;PROFILE&#039;&#039;&#039; representing its sound and dynamic behavior.<br /> <br /> Kemper&#039;s profiling technology has evolved considerably since its introduction.<br /> <br /> The current ecosystem includes:<br /> <br /> * Traditional PROFILING<br /> * Direct and cabinet-inclusive PROFILEs<br /> * Merged PROFILEs<br /> * Liquid Profiling<br /> * PROFILING 2.0<br /> * Rig Manager<br /> * Rig Exchange<br /> * PROFILER hardware<br /> <br /> == PROFILEs and Rigs ==<br /> <br /> In Kemper terminology, a &#039;&#039;&#039;PROFILE&#039;&#039;&#039; represents the captured amplifier or signal-chain behavior.<br /> <br /> A &#039;&#039;&#039;Rig&#039;&#039;&#039; is the larger playable configuration containing the PROFILE together with other PROFILER settings.<br /> <br /> A Rig can include:<br /> <br /> * Amplifier PROFILE<br /> * Cabinet<br /> * Effects<br /> * Input and output configuration<br /> * Rig-specific settings<br /> * Performance-related parameters<br /> <br /> The distinction is useful when discussing shared content.<br /> <br /> Two Rigs can use the same underlying PROFILE while containing different effects or other settings.<br /> <br /> == Basic profiling process ==<br /> <br /> A conventional Kemper profiling session compares a real reference amplifier with the PROFILER&#039;s own test signals.<br /> <br /> A simplified signal path is:<br /> <br /> &#039;&#039;&#039;PROFILER test signal → reference amplifier → return path → PROFILER&#039;&#039;&#039;<br /> <br /> For a miked amplifier:<br /> <br /> &#039;&#039;&#039;PROFILER → amplifier → cabinet → microphone → PROFILER Return Input&#039;&#039;&#039;<br /> <br /> The PROFILER sends measurement signals through the reference equipment and analyzes the returned response.<br /> <br /> After the initial PROFILE is created, the user can compare the reference amplifier and PROFILE directly.<br /> <br /> == Reference amplifier ==<br /> <br /> Kemper uses the term &#039;&#039;&#039;Reference Amplifier&#039;&#039;&#039; for the physical amplifier or rig being profiled.<br /> <br /> The objective is to reproduce the behavior and sound of that specific setup.<br /> <br /> The Reference Amplifier can be configured using the settings chosen by the user before profiling begins.<br /> <br /> These settings are therefore part of the captured operating point.<br /> <br /> == Studio PROFILE ==<br /> <br /> A &#039;&#039;&#039;Studio PROFILE&#039;&#039;&#039; is created using the complete miked amplifier chain.<br /> <br /> A typical path is:<br /> <br /> &#039;&#039;&#039;PROFILER → amplifier → speaker cabinet → microphone → microphone preamp/return → PROFILER&#039;&#039;&#039;<br /> <br /> The resulting PROFILE contains the combined response of the amplifier, cabinet, microphone, and associated return path.<br /> <br /> This is similar in concept to other capture systems that create a complete amp/cab/microphone model.<br /> <br /> The microphone and its placement are therefore part of the resulting sound.<br /> <br /> == Direct Amp PROFILE ==<br /> <br /> A &#039;&#039;&#039;Direct Amp PROFILE&#039;&#039;&#039; captures an amplifier without the acoustic cabinet and microphone response.<br /> <br /> The amplifier&#039;s speaker output is measured using equipment designed to obtain an appropriate signal while maintaining a proper speaker load.<br /> <br /> The resulting PROFILE can then be combined with:<br /> <br /> * A Kemper Cabinet<br /> * A cabinet derived from another PROFILE<br /> * An impulse response converted for use with the PROFILER<br /> * A physical guitar cabinet<br /> <br /> Direct Amp PROFILEs provide greater flexibility when the user wants to change the cabinet independently.<br /> <br /> &#039;&#039;&#039;Never connect an amplifier speaker output directly to an ordinary line or instrument input.&#039;&#039;&#039;<br /> <br /> Appropriate load and signal-level equipment must be used.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Merged PROFILE ==<br /> <br /> A &#039;&#039;&#039;Merged PROFILE&#039;&#039;&#039; is created by combining information from:<br /> <br /> * A Studio PROFILE<br /> * A corresponding Direct Amp PROFILE<br /> <br /> of the same amplifier configuration.<br /> <br /> The PROFILER uses these measurements to separate the amplifier portion from the cabinet portion more accurately.<br /> <br /> This allows the amplifier and cabinet components to be used independently.<br /> <br /> Merged PROFILEs are particularly useful when the same amplifier sound must work both:<br /> <br /> * Direct to a full-range system with cabinet simulation<br /> * Through a physical guitar cabinet with the virtual cabinet disabled<br /> <br /> == Cabinet separation ==<br /> <br /> Kemper PROFILEs can provide separate Amp and Cabinet sections.<br /> <br /> With ordinary Studio PROFILEs, the division between amplifier and cabinet is estimated by the PROFILER.<br /> <br /> A Merged PROFILE provides additional measurement information intended to improve that separation.<br /> <br /> This distinction matters when exchanging cabinets or using a physical guitar cabinet.<br /> <br /> See [[Capture Types]] and [[Cabinets and IRs]].<br /> <br /> == Refining ==<br /> <br /> Traditional Kemper PROFILING includes a &#039;&#039;&#039;Refine&#039;&#039;&#039; process.<br /> <br /> After the initial PROFILE is created, the player can send guitar material through the reference setup while the PROFILER further analyzes differences between the reference amplifier and the PROFILE.<br /> <br /> Refining can improve the result for some rigs.<br /> <br /> It should be treated as part of the profiling workflow rather than as a substitute for correct routing and levels.<br /> <br /> The current PROFILER software may automate or alter aspects of the process depending on the profiling mode and software generation.<br /> <br /> Consult the current Kemper documentation for the exact procedure.<br /> <br /> == A/B comparison ==<br /> <br /> Kemper places particular emphasis on direct comparison between the Reference Amplifier and the resulting PROFILE.<br /> <br /> During profiling, the user can switch between:<br /> <br /> * Reference Amp<br /> * PROFILER<br /> <br /> This allows the creator to evaluate the result before saving it.<br /> <br /> Listen for differences in:<br /> <br /> * Gain<br /> * Frequency balance<br /> * Pick attack<br /> * Dynamics<br /> * Compression<br /> * Low-frequency response<br /> * High-frequency response<br /> * Sustain<br /> <br /> Level matching remains important when making subjective comparisons.<br /> <br /> == Snapshot behavior ==<br /> <br /> Traditional PROFILEs originate from a particular configuration of the Reference Amplifier.<br /> <br /> The physical amplifier may have been profiled with specific:<br /> <br /> * Gain<br /> * Bass<br /> * Middle<br /> * Treble<br /> * Presence<br /> * Master volume<br /> * Channel<br /> * Switch settings<br /> <br /> The PROFILER provides controls that allow the resulting PROFILE to be modified substantially.<br /> <br /> However, the original profiling system did not reproduce the complete physical behavior of every original amplifier control across its entire range.<br /> <br /> This limitation led to the development of &#039;&#039;&#039;Liquid Profiling&#039;&#039;&#039;.<br /> <br /> == Liquid Profiling ==<br /> <br /> &#039;&#039;&#039;Liquid Profiling&#039;&#039;&#039; extends Kemper&#039;s profiling system by combining the captured PROFILE with modeled behavior of the original amplifier&#039;s controls.<br /> <br /> Kemper provides &#039;&#039;&#039;Amp Models&#039;&#039;&#039; representing the gain and tone-stack behavior of specific amplifier designs.<br /> <br /> When the appropriate Amp Model is associated with a PROFILE, controls such as:<br /> <br /> * Gain<br /> * Bass<br /> * Middle<br /> * Treble<br /> <br /> can behave more like the corresponding controls of the original amplifier.<br /> <br /> The objective is to move beyond treating the PROFILE solely as a fixed operating-point snapshot.<br /> <br /> == Creating a Liquid Profile ==<br /> <br /> For an authentic Liquid Profile workflow, the creator selects the appropriate Kemper Amp Model and records information about the actual control settings used on the Reference Amplifier.<br /> <br /> The PROFILER can then relate the captured operating point to the modeled control behavior.<br /> <br /> For best correspondence, the Amp Model should match the circuit behavior of the amplifier being profiled.<br /> <br /> The original amplifier settings are therefore valuable information rather than merely descriptive metadata.<br /> <br /> == Retrofitting Liquid Profiling ==<br /> <br /> Existing PROFILEs can also be associated with a Liquid Profiling Amp Model.<br /> <br /> This allows older PROFILEs to benefit from modeled tone-stack and gain behavior without physically profiling the amplifier again.<br /> <br /> The result is most meaningful when the original amplifier settings and circuit type are known.<br /> <br /> Applying an unrelated Amp Model is also possible as a creative tool, but should not be confused with reconstructing the authentic controls of the original amplifier.<br /> <br /> == Generic Amp Model ==<br /> <br /> Kemper also provides a generic amplifier-control behavior.<br /> <br /> When no specific Liquid Profiling Amp Model is selected, the PROFILE can continue to use the more traditional generic PROFILER controls.<br /> <br /> This maintains compatibility with the enormous library of PROFILEs created before Liquid Profiling existed.<br /> <br /> == PROFILING 2.0 ==<br /> <br /> &#039;&#039;&#039;PROFILING 2.0&#039;&#039;&#039; is Kemper&#039;s current-generation profiling technology.<br /> <br /> Introduced for the PROFILER MK 2 generation and current PROFILER OS, it increases the resolution of the profiling process and extends the measurement of speaker and cabinet resonance.<br /> <br /> Kemper describes PROFILING 2.0 as using more than &#039;&#039;&#039;100,000 frequency points&#039;&#039;&#039; in its high-resolution profiling process.<br /> <br /> It also integrates Liquid Profiling into the current profiling workflow.<br /> <br /> == PROFILING 2.0 and MK 2 ==<br /> <br /> PROFILER MK 2 hardware can reproduce PROFILING 2.0 content at its full high-resolution capability.<br /> <br /> The current MK 2 family includes updated versions of:<br /> <br /> * PROFILER Head<br /> * PROFILER Rack<br /> * PROFILER Stage<br /> * PROFILER Remote<br /> <br /> The MK 2 platform also provides increased processing resources for effects and other functions.<br /> <br /> == PROFILING 2.0 and MK 1 ==<br /> <br /> PROFILING 2.0 content does not make older PROFILER hardware obsolete.<br /> <br /> With PROFILER OS 14, MK 1 PROFILER hardware can load and play PROFILING 2.0 Rigs at standard resolution.<br /> <br /> The same Rig can therefore remain compatible across PROFILER generations while newer hardware uses the higher-resolution representation.<br /> <br /> This distinction is important when discussing compatibility:<br /> <br /> &#039;&#039;&#039;PROFILING 2.0 Rig on MK 2 → high-resolution playback&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;PROFILING 2.0 Rig on MK 1 with compatible OS → standard-resolution playback&#039;&#039;&#039;<br /> <br /> The content does not require a completely separate library for each hardware generation.<br /> <br /> == PROFILER Player ==<br /> <br /> The &#039;&#039;&#039;PROFILER Player&#039;&#039;&#039; is Kemper&#039;s compact playback-oriented hardware platform.<br /> <br /> It can load PROFILEs and Rigs from the Kemper ecosystem.<br /> <br /> Its capabilities differ from the larger PROFILER units and can also depend on the installed feature level and software version.<br /> <br /> The Player should therefore not automatically be assumed to provide every profiling or editing function available on a Head, Rack, or Stage.<br /> <br /> For example, some operations related to creating or retrofitting Liquid Profiles may require other PROFILER models.<br /> <br /> == Effects ==<br /> <br /> The PROFILER is also a complete effects processor.<br /> <br /> A Rig can contain processing before and after the amplifier section.<br /> <br /> Available effect categories include processing such as:<br /> <br /> * Drives<br /> * Distortion<br /> * Fuzz<br /> * Compression<br /> * EQ<br /> * Modulation<br /> * Pitch effects<br /> * Delay<br /> * Reverb<br /> <br /> These effects are separate from the PROFILE itself.<br /> <br /> When evaluating a PROFILE, distinguish between the captured amplifier behavior and additional effects contained in the Rig.<br /> <br /> == Cabinet section ==<br /> <br /> The PROFILER provides a separate Cabinet section.<br /> <br /> Cabinets can be exchanged between compatible Rigs.<br /> <br /> A user can therefore combine an amplifier PROFILE with a different cabinet.<br /> <br /> Cabinet processing can also be disabled when using an appropriate physical guitar cabinet.<br /> <br /> This flexibility is one of the reasons it is useful to understand whether a PROFILE is:<br /> <br /> * Studio<br /> * Direct Amp<br /> * Merged<br /> <br /> == Impulse responses ==<br /> <br /> Third-party cabinet impulse responses can be used within the Kemper ecosystem after appropriate import/conversion through supported Kemper software.<br /> <br /> An IR represents a cabinet/microphone response rather than an amplifier PROFILE.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Rig Manager ==<br /> <br /> &#039;&#039;&#039;Rig Manager&#039;&#039;&#039; is Kemper&#039;s free software for managing PROFILER content.<br /> <br /> It is available for desktop and supported mobile platforms.<br /> <br /> Depending on platform and connected hardware, Rig Manager provides functions including:<br /> <br /> * Browsing Rigs<br /> * Editing Rig parameters<br /> * Organizing libraries<br /> * Managing Performances<br /> * Managing presets<br /> * Accessing Rig Exchange<br /> * Editing effects<br /> * Managing connected PROFILER hardware<br /> <br /> Rig Manager is an important part of the Kemper ecosystem even though PROFILE playback itself occurs on PROFILER hardware.<br /> <br /> == Rig Exchange ==<br /> <br /> &#039;&#039;&#039;Rig Exchange&#039;&#039;&#039; is Kemper&#039;s community library for sharing Rigs.<br /> <br /> It contains thousands of user-created Rigs representing a wide range of amplifiers and equipment.<br /> <br /> Kemper currently states that Rig Exchange contains more than &#039;&#039;&#039;20,000 Rigs&#039;&#039;&#039;.<br /> <br /> Because the library continues to change, exact counts should be treated as time-sensitive.<br /> <br /> Useful Rig information can include:<br /> <br /> * Author<br /> * Amplifier<br /> * Cabinet<br /> * Gain<br /> * Tags<br /> * Comments<br /> * Other identifying information<br /> <br /> As with any community capture library, good documentation greatly increases the long-term value of the content.<br /> <br /> == Commercial PROFILEs ==<br /> <br /> In addition to free Rig Exchange content, a substantial commercial ecosystem exists around Kemper PROFILEs.<br /> <br /> Third-party creators sell collections of PROFILEs representing:<br /> <br /> * Amplifiers<br /> * Cabinets<br /> * Complete rigs<br /> * Artist tones<br /> * Studio signal chains<br /> <br /> The quality, documentation, and capture methodology of third-party PROFILEs vary by creator.<br /> <br /> A commercial PROFILE should not automatically be assumed to be more or less accurate than a community-created PROFILE.<br /> <br /> Evaluate the result on its own merits.<br /> <br /> == Sharing between PROFILER models ==<br /> <br /> PROFILEs and Rigs are designed to move between compatible PROFILER hardware.<br /> <br /> This distinguishes Kemper from systems in which a capture is permanently tied to the individual unit that created it.<br /> <br /> Compatibility can nevertheless depend on:<br /> <br /> * PROFILER OS version<br /> * Hardware generation<br /> * Feature availability<br /> * Rig features<br /> <br /> Current documentation should be consulted when using newer content on older hardware.<br /> <br /> == Kemper and NAM ==<br /> <br /> Kemper Profiling and [[Neural Amp Modeler]] both reproduce the behavior of physical audio equipment, but they use different technologies and ecosystems.<br /> <br /> A Kemper PROFILE is not a &#039;&#039;.nam&#039;&#039; file.<br /> <br /> The systems differ in:<br /> <br /> * Capture process<br /> * Model representation<br /> * Playback implementation<br /> * Hardware architecture<br /> * Model distribution<br /> * Editing<br /> * Control modeling<br /> <br /> NAM provides an open-source model and playback ecosystem.<br /> <br /> Kemper PROFILING is a proprietary Kemper technology implemented within the PROFILER ecosystem.<br /> <br /> Neither distinction alone determines which system will sound better for a particular capture.<br /> <br /> == Kemper and TONEX ==<br /> <br /> Kemper PROFILING and [[TONEX]] both provide integrated capture and playback ecosystems.<br /> <br /> Both can represent:<br /> <br /> * Amplifiers<br /> * Cabinets<br /> * Complete rigs<br /> <br /> Their capture methods, model formats, editing systems, and hardware implementations are different.<br /> <br /> Models are not directly interchangeable between the systems.<br /> <br /> == Profiling versus conventional modeling ==<br /> <br /> Traditional algorithmic amplifier modeling generally begins with a developer creating a model intended to represent a particular amplifier design.<br /> <br /> Kemper takes a different approach.<br /> <br /> The user measures a specific physical rig and creates a PROFILE from it.<br /> <br /> Liquid Profiling adds a hybrid element by combining that measured PROFILE with modeled control behavior based on the original amplifier circuit.<br /> <br /> The modern Kemper system therefore combines:<br /> <br /> &#039;&#039;&#039;captured amplifier behavior + modeled amplifier control behavior&#039;&#039;&#039;<br /> <br /> This distinction is important when comparing current Kemper technology with both conventional modeling and snapshot-oriented capture systems.<br /> <br /> == Documentation ==<br /> <br /> When creating PROFILEs for other users, useful information includes:<br /> <br /> * Amplifier manufacturer<br /> * Amplifier model<br /> * Channel<br /> * Amplifier settings<br /> * PROFILE type<br /> * Cabinet<br /> * Speaker<br /> * Microphone<br /> * Microphone position<br /> * Direct/load equipment<br /> * Liquid Amp Model<br /> * Profiling generation<br /> * Creator notes<br /> <br /> For Liquid Profiles, preserving the original amplifier control settings is particularly useful.<br /> <br /> == Troubleshooting ==<br /> <br /> If a PROFILE differs significantly from the Reference Amplifier, check:<br /> <br /> * Profiling signal path<br /> * Return level<br /> * Reference amplifier configuration<br /> * Microphone path<br /> * Load/direct-output configuration<br /> * Cabinet configuration<br /> * Effects<br /> * A/B comparison level<br /> * Liquid Amp Model selection<br /> * PROFILER OS version<br /> <br /> For Direct Amp profiling, verify that the amplifier is connected to an appropriate load and that the signal reaching the PROFILER is suitable for its input.<br /> <br /> For playback problems, also determine whether the issue is in the PROFILE itself or elsewhere in the Rig.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Official resources ==<br /> <br /> * [https://www.kemper-amps.com/ Kemper official website]<br /> * [https://www.kemper-amps.com/profiler/overview KEMPER PROFILER overview]<br /> * [https://www.kemper-amps.com/start PROFILER tutorials and getting-started resources]<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Neural Amp Modeler]]<br /> * [[TONEX]]<br /> * [[Capture Types]]<br /> * [[Creating a Capture]]<br /> * [[Capture Signal Chain]]<br /> * [[Reamping for Capture]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss Kemper PROFILEs, Liquid Profiling, and PROFILER workflows on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Kemper]]<br /> [[Category:Capture technology]]<br /> [[Category:Commercial capture systems]]</div> Tue, 01 Sep 2026 06:05:02 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Kemper_Profiling https://namforum.com/wiki/index.php?title=TONEX&diff=16&oldid=0 https://namforum.com/wiki/index.php?title=TONEX&diff=16&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= TONEX =<br /> <br /> &#039;&#039;&#039;TONEX&#039;&#039;&#039; is a capture, modeling, playback, and tone-sharing ecosystem developed by IK Multimedia for guitar and bass equipment.<br /> <br /> TONEX uses IK Multimedia&#039;s &#039;&#039;&#039;AI Machine Modeling&#039;&#039;&#039; technology to create digital models called &#039;&#039;&#039;Tone Models&#039;&#039;&#039; from real amplifiers, pedals, cabinets, and complete signal chains.<br /> <br /> Tone Models can be created using [[#TONEX Modeler|TONEX Modeler]], played in TONEX software, transferred to compatible TONEX hardware, integrated with AmpliTube, and shared through IK Multimedia&#039;s ToneNET service.<br /> <br /> == Overview ==<br /> <br /> A TONEX workflow can include four distinct activities:<br /> <br /> # Capturing real equipment<br /> # Training a Tone Model<br /> # Playing and editing Tone Models<br /> # Sharing and transferring models and presets<br /> <br /> The TONEX ecosystem includes dedicated software for these functions as well as hardware designed to run Tone Models without a computer.<br /> <br /> == AI Machine Modeling ==<br /> <br /> IK Multimedia refers to the technology underlying TONEX as &#039;&#039;&#039;AI Machine Modeling&#039;&#039;&#039;.<br /> <br /> The system analyzes the relationship between a known input signal and the response produced by the equipment being modeled.<br /> <br /> A simplified capture path is:<br /> <br /> &#039;&#039;&#039;TONEX capture signal → audio interface → reamp/capture path → device under test → audio interface → TONEX Modeler&#039;&#039;&#039;<br /> <br /> The captured data is then processed to create a Tone Model representing the behavior of the equipment.<br /> <br /> Like other data-driven capture technologies, TONEX differs from conventional component-by-component algorithmic amplifier modeling.<br /> <br /> The objective is to learn the behavior of a particular physical rig or device at the configuration used during capture.<br /> <br /> == Tone Models ==<br /> <br /> A &#039;&#039;&#039;Tone Model&#039;&#039;&#039; is the modeled representation created by TONEX.<br /> <br /> Tone Models can represent different portions of a signal chain, including:<br /> <br /> * Amplifiers<br /> * Amplifiers and cabinets<br /> * Combo amplifiers<br /> * Pedals<br /> * Complete rigs<br /> <br /> Supported pedal targets include nonlinear devices such as:<br /> <br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> * Boost<br /> * EQ<br /> <br /> Multiple pieces of equipment can also be captured as part of a combined signal chain.<br /> <br /> See [[Capture Types]].<br /> <br /> == Snapshot nature of Tone Models ==<br /> <br /> A Tone Model represents equipment in the configuration used when it was captured.<br /> <br /> For example, an amplifier captured with particular settings for gain, tone controls, channel, switches, and master volume represents the behavior measured at that configuration.<br /> <br /> Controls available during TONEX playback can modify the resulting sound, but they should not automatically be interpreted as recreating the complete physical control circuitry of the original amplifier.<br /> <br /> Different operating points can therefore be represented by separate Tone Models.<br /> <br /> == TONEX Modeler ==<br /> <br /> &#039;&#039;&#039;TONEX Modeler&#039;&#039;&#039; is IK Multimedia&#039;s dedicated application for capturing and training Tone Models.<br /> <br /> Capture and training were originally integrated more directly into the TONEX software workflow. The current architecture separates model creation into the dedicated TONEX Modeler application.<br /> <br /> TONEX Modeler guides the user through:<br /> <br /> * Instrument and Tone Model type<br /> * Audio configuration<br /> * Rig connections<br /> * Input level<br /> * Send level<br /> * Return level<br /> * Capture<br /> * Training<br /> * Model comparison<br /> * Metadata<br /> * Finalization<br /> <br /> The resulting Tone Model can then be used throughout the TONEX ecosystem.<br /> <br /> == Capture and training are separate stages ==<br /> <br /> The current TONEX Modeler separates the physical capture from neural-model training.<br /> <br /> This has an important practical advantage.<br /> <br /> The physical equipment can be captured once and the resulting capture data can be trained later.<br /> <br /> A user can therefore:<br /> <br /> # Connect and configure the physical rig.<br /> # Record several captures.<br /> # Disconnect or change the equipment.<br /> # Train the captured data afterward.<br /> <br /> Training can also be performed on a different computer.<br /> <br /> This is useful when the machine connected to the studio equipment is not the fastest available machine for neural-network training.<br /> <br /> == Portable capture files ==<br /> <br /> The current TONEX Modeler stores capture data as portable WAV files with embedded metadata.<br /> <br /> This allows capture sessions to be:<br /> <br /> * Archived<br /> * Moved between computers<br /> * Trained later<br /> * Batch processed<br /> * Preserved independently of the finished Tone Model<br /> <br /> Preserving original capture data can be valuable as training technology evolves.<br /> <br /> The physical equipment does not necessarily need to be recaptured merely because a newer training implementation becomes available.<br /> <br /> == Batch processing ==<br /> <br /> TONEX Modeler supports batch workflows.<br /> <br /> Multiple rigs or settings can be captured during one session and trained afterward.<br /> <br /> This is particularly useful when creating collections containing:<br /> <br /> * Multiple amplifier channels<br /> * Several gain settings<br /> * Different pedals<br /> * Different cabinet configurations<br /> * Multiple complete rigs<br /> <br /> The time-consuming physical setup can therefore be separated from the computational training process.<br /> <br /> == Training ==<br /> <br /> TONEX Modeler analyzes the captured response and creates a Tone Model.<br /> <br /> The current Modeler is optimized for modern NVIDIA GPUs and Apple Silicon systems while also supporting CPU-based training.<br /> <br /> IK Multimedia states that the current Modeler substantially reduces training time compared with its previous training workflow.<br /> <br /> Actual training time depends on factors including:<br /> <br /> * Computer hardware<br /> * Capture complexity<br /> * Training process<br /> * Model requirements<br /> <br /> The Modeler can automatically stop training when its analysis determines that the model has reached its optimum result.<br /> <br /> == Comparing the model with the rig ==<br /> <br /> TONEX Modeler provides a final comparison stage where the resulting Tone Model can be evaluated against the original captured rig.<br /> <br /> This is an important part of the process.<br /> <br /> Successful completion of training does not eliminate the need to listen critically.<br /> <br /> Compare:<br /> <br /> * Gain<br /> * Distortion<br /> * Dynamics<br /> * Attack<br /> * Frequency response<br /> * Sustain<br /> * Low-frequency behavior<br /> * High-frequency behavior<br /> <br /> Where possible, comparisons should be level matched.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Phase alignment ==<br /> <br /> The current TONEX Modeler is designed to create phase-aligned models.<br /> <br /> This is particularly useful when Tone Models are combined with parallel signals.<br /> <br /> Examples include:<br /> <br /> * Bass DI combined with a modeled amplifier<br /> * Stereo rigs<br /> * Parallel processing<br /> <br /> Phase and latency differences can cause cancellations when multiple versions of related signals are mixed.<br /> <br /> Maintaining predictable alignment therefore has practical value beyond ordinary single-path guitar playback.<br /> <br /> == Capture levels ==<br /> <br /> Correct levels are important when creating Tone Models.<br /> <br /> TONEX Modeler includes dedicated stages for:<br /> <br /> * Input level<br /> * Send level<br /> * Return level<br /> <br /> These should be treated as distinct parts of the capture chain.<br /> <br /> A level problem at one stage should not automatically be corrected by changing an unrelated stage.<br /> <br /> For nonlinear equipment, the level presented to the device affects the behavior being captured.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Reamping and TONEX Capture ==<br /> <br /> Capturing conventional guitar amplifiers and pedals normally requires an appropriate connection between the audio interface and the device being modeled.<br /> <br /> IK Multimedia offers the &#039;&#039;&#039;TONEX Capture&#039;&#039;&#039; accessory for this purpose.<br /> <br /> TONEX Capture provides the signal-routing and level-conversion functions needed to connect audio equipment to amplifiers and pedals during the modeling process.<br /> <br /> A TONEX Capture device is not required in principle if another correctly configured capture/reamp solution performs the necessary functions.<br /> <br /> The important requirement is a suitable and correctly configured signal path.<br /> <br /> See [[Reamping for Capture]].<br /> <br /> == Capturing amplifiers ==<br /> <br /> An amplifier can be captured without its physical cabinet response.<br /> <br /> This produces a model that can later be combined with:<br /> <br /> * TONEX cabinet processing<br /> * VIR cabinet processing<br /> * A custom IR<br /> * A physical guitar cabinet<br /> <br /> This provides flexibility because the amplifier and cabinet portions of the signal chain can be changed independently.<br /> <br /> == Capturing complete rigs ==<br /> <br /> TONEX can also model a signal chain containing the cabinet and microphone.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Capture signal → amplifier → cabinet → microphone → interface&#039;&#039;&#039;<br /> <br /> The resulting Tone Model includes the characteristics of the measured return path.<br /> <br /> Depending on the capture configuration, this can include:<br /> <br /> * Amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Microphone preamp<br /> <br /> Additional cabinet processing may therefore be unnecessary.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Pedal captures ==<br /> <br /> TONEX can create Tone Models of supported nonlinear pedals.<br /> <br /> Typical examples include:<br /> <br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> * Boost<br /> * EQ<br /> <br /> A pedal Tone Model can then be combined with an amplifier Tone Model or other processing during playback.<br /> <br /> Capturing the pedal independently provides more flexibility than capturing the pedal and amplifier together.<br /> <br /> Conversely, capturing both together may preserve the particular interaction of that combination.<br /> <br /> Neither approach is universally preferable.<br /> <br /> == TONEX Player ==<br /> <br /> &#039;&#039;&#039;TONEX Player&#039;&#039;&#039; is the primary desktop environment for discovering, organizing, editing, playing, and recording with Tone Models and presets.<br /> <br /> It can operate as:<br /> <br /> * Standalone software<br /> * A plugin in a DAW<br /> <br /> The current TONEX 2.0 Player introduces a redesigned workflow with:<br /> <br /> * Smart Search<br /> * Dedicated gear browsers<br /> * Custom folders<br /> * Favorites<br /> * Recently used content<br /> * Enhanced ToneNET integration<br /> * Expanded signal-chain editing<br /> <br /> As of September 2026, TONEX 2.0 Player is available as a &#039;&#039;&#039;public beta&#039;&#039;&#039; to existing TONEX users.<br /> <br /> Software status should be checked against IK Multimedia&#039;s current documentation because this will change when TONEX 2.0 reaches its production release.<br /> <br /> == Tone Model versus preset ==<br /> <br /> A &#039;&#039;&#039;Tone Model&#039;&#039;&#039; and a &#039;&#039;&#039;TONEX preset&#039;&#039;&#039; are not the same thing.<br /> <br /> The Tone Model represents the captured equipment.<br /> <br /> A preset can contain the Tone Model together with additional TONEX configuration and processing.<br /> <br /> A preset may therefore include:<br /> <br /> * Tone Model<br /> * Cabinet configuration<br /> * EQ<br /> * Compression<br /> * Noise gate<br /> * Modulation<br /> * Delay<br /> * Reverb<br /> * Other TONEX settings<br /> <br /> This distinction is important when sharing tones.<br /> <br /> Two presets using the same underlying Tone Model can sound substantially different.<br /> <br /> == Tone shaping ==<br /> <br /> TONEX provides additional processing around the captured Tone Model.<br /> <br /> Depending on the current software or hardware implementation, this can include:<br /> <br /> * EQ<br /> * Compression<br /> * Noise gate<br /> * Modulation<br /> * Delay<br /> * Reverb<br /> * Cabinet processing<br /> <br /> These effects are not necessarily part of the original captured model.<br /> <br /> They form part of the playback signal chain.<br /> <br /> When evaluating the accuracy of a Tone Model, distinguish the model itself from additional processing in the preset.<br /> <br /> == VIR cabinet technology ==<br /> <br /> TONEX includes IK Multimedia&#039;s &#039;&#039;&#039;VIR&#039;&#039;&#039; cabinet technology.<br /> <br /> VIR provides adjustable cabinet and microphone processing that can be combined with amplifier Tone Models.<br /> <br /> TONEX also supports custom impulse responses.<br /> <br /> This allows a direct amplifier Tone Model to be paired with:<br /> <br /> * VIR cabinets<br /> * Custom IRs<br /> * Other appropriate cabinet processing<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Custom IRs ==<br /> <br /> TONEX software and compatible TONEX hardware can load custom cabinet IRs.<br /> <br /> This allows users to combine a direct Tone Model with cabinet responses created outside the TONEX ecosystem.<br /> <br /> Before adding an IR, determine whether the Tone Model already contains a cabinet.<br /> <br /> Applying another cabinet response to a complete amp/cab capture may produce unintended double cabinet processing.<br /> <br /> == ToneNET ==<br /> <br /> &#039;&#039;&#039;ToneNET&#039;&#039;&#039; is IK Multimedia&#039;s online community platform for sharing and discovering tones.<br /> <br /> TONEX integration allows users to browse, search, audition, download, and share Tone Models and presets.<br /> <br /> The current TONEX ecosystem provides access to tens of thousands of community-created Tone Models and presets.<br /> <br /> ToneNET also distributes commercial and curated content including Premium, Signature, and Tone Partner collections.<br /> <br /> Because the library continually changes, exact model counts should be treated as time-sensitive rather than permanent specifications.<br /> <br /> == ToneNET model sharing ==<br /> <br /> A creator can upload a Tone Model to ToneNET so other TONEX users can discover and use it.<br /> <br /> Useful model information includes:<br /> <br /> * Equipment manufacturer<br /> * Equipment model<br /> * Capture type<br /> * Amplifier settings<br /> * Pedal settings<br /> * Cabinet<br /> * Microphone<br /> * Capture notes<br /> <br /> Good metadata is especially valuable in a large community library.<br /> <br /> A model called only &quot;Marshall&quot; or &quot;High Gain&quot; provides considerably less useful information than a documented capture.<br /> <br /> == TONEX Editor ==<br /> <br /> &#039;&#039;&#039;TONEX Editor&#039;&#039;&#039; is a standalone application for managing and editing compatible TONEX hardware.<br /> <br /> It provides access to:<br /> <br /> * Hardware presets<br /> * Tone Models<br /> * IRs<br /> * Computer library<br /> * ToneNET<br /> * Global hardware settings<br /> <br /> Tone Models and presets can be transferred between the computer and supported TONEX hardware.<br /> <br /> The Editor also allows models stored on the computer to be auditioned through connected hardware.<br /> <br /> == Hardware ecosystem ==<br /> <br /> TONEX Tone Models can be used in several IK Multimedia hardware products.<br /> <br /> The current ecosystem includes products such as:<br /> <br /> * TONEX Pedal<br /> * TONEX ONE<br /> * TONEX ONE+<br /> * TONEX Plug<br /> <br /> These products serve different roles and have different storage, control, connectivity, and preset capabilities.<br /> <br /> The important common feature is the ability to use TONEX Tone Models outside the desktop computer environment.<br /> <br /> == TONEX Pedal ==<br /> <br /> &#039;&#039;&#039;TONEX Pedal&#039;&#039;&#039; is IK Multimedia&#039;s larger dedicated TONEX floor pedal.<br /> <br /> It provides hardware playback of Tone Models together with preset management, cabinet processing, effects, MIDI, USB audio, and live-performance controls.<br /> <br /> Tone Models can be transferred to the pedal using TONEX software.<br /> <br /> == TONEX ONE ==<br /> <br /> &#039;&#039;&#039;TONEX ONE&#039;&#039;&#039; is a compact TONEX hardware pedal.<br /> <br /> It can store Tone Models and presets for pedalboard use and includes cabinet processing and effects.<br /> <br /> Despite its small size, it can represent:<br /> <br /> * Pedals<br /> * Amplifiers<br /> * Complete rigs<br /> <br /> and can load custom IRs.<br /> <br /> == TONEX ONE+ ==<br /> <br /> &#039;&#039;&#039;TONEX ONE+&#039;&#039;&#039; extends the compact TONEX hardware concept with additional control and connectivity.<br /> <br /> It includes MIDI capability and works with IK Multimedia&#039;s TONEX Control mobile application for wireless editing and tone management.<br /> <br /> Hardware capabilities should be checked against current IK specifications because the TONEX hardware family continues to evolve.<br /> <br /> == TONEX Plug ==<br /> <br /> &#039;&#039;&#039;TONEX Plug&#039;&#039;&#039; is a portable TONEX headphone amplifier and USB audio interface.<br /> <br /> It stores TONEX presets and can be controlled wirelessly using the TONEX Control mobile application.<br /> <br /> Its inclusion in the ecosystem illustrates the portability of Tone Models across different forms of playback hardware.<br /> <br /> == TONEX Control ==<br /> <br /> &#039;&#039;&#039;TONEX Control&#039;&#039;&#039; is IK Multimedia&#039;s mobile application for compatible TONEX hardware.<br /> <br /> It provides functions including:<br /> <br /> * Preset editing<br /> * Tone Model browsing<br /> * ToneNET access<br /> * Tone transfer<br /> * Preset management<br /> * Hardware control<br /> <br /> The application is available for supported iOS and Android devices.<br /> <br /> == Integration with AmpliTube ==<br /> <br /> TONEX also integrates with IK Multimedia&#039;s &#039;&#039;&#039;AmpliTube&#039;&#039;&#039; ecosystem.<br /> <br /> Tone Models can be used as part of larger virtual guitar and bass signal chains containing other modeled equipment and effects.<br /> <br /> This allows capture-based Tone Models and conventional software effects/modeling to coexist in the same production environment.<br /> <br /> == TONEX and NAM ==<br /> <br /> TONEX and [[Neural Amp Modeler]] both use machine-learning techniques to reproduce the behavior of real audio equipment, but they are separate technologies.<br /> <br /> A TONEX Tone Model is not a &#039;&#039;.nam&#039;&#039; model.<br /> <br /> The systems have different:<br /> <br /> * Model formats<br /> * Training implementations<br /> * Playback engines<br /> * Software ecosystems<br /> * Hardware ecosystems<br /> * Distribution systems<br /> <br /> A model created for one system should not be assumed to run directly in the other.<br /> <br /> They can nevertheless be used to model the same physical equipment and can be compared as alternative capture technologies.<br /> <br /> == Closed and open ecosystems ==<br /> <br /> A major architectural difference between TONEX and NAM is ecosystem control.<br /> <br /> NAM provides an open model format and open-source playback implementation that third-party developers can implement independently.<br /> <br /> TONEX is an IK Multimedia technology whose model creation, playback, hardware integration, and ToneNET distribution are built around the TONEX ecosystem.<br /> <br /> This does not by itself determine modeling quality.<br /> <br /> It describes a significant difference in how the two technologies are developed and distributed.<br /> <br /> == Preserving captures ==<br /> <br /> The current TONEX Modeler&#039;s ability to preserve captured responses separately from final Tone Models is useful for long-term model creation.<br /> <br /> A good archive can include:<br /> <br /> * Original capture WAV<br /> * Equipment settings<br /> * Capture signal chain<br /> * Cabinet and microphone information<br /> * Interface settings<br /> * Modeler version<br /> * Finished Tone Model<br /> * Notes<br /> <br /> Preserving the original measurement can make future retraining or investigation possible without reconnecting the original equipment.<br /> <br /> == Troubleshooting ==<br /> <br /> If a TONEX capture does not reproduce the expected sound, investigate the complete capture chain.<br /> <br /> Check:<br /> <br /> * Input level<br /> * Send level<br /> * Return level<br /> * Interface routing<br /> * Reamp/capture path<br /> * Clipping<br /> * Unwanted processing<br /> * Cabinet configuration<br /> * Model comparison<br /> * Playback preset<br /> <br /> If a Tone Model sounds correct in TONEX software but different in hardware, also check:<br /> <br /> * Preset processing<br /> * Cabinet/IR settings<br /> * Input trim<br /> * Output configuration<br /> * Global cabinet bypass<br /> * Hardware firmware<br /> * Monitoring system<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Official resources ==<br /> <br /> * [https://www.ikmultimedia.com/products/tonex/ TONEX official website]<br /> * [https://www.ikmultimedia.com/products/tonex/index.php?p=modeler TONEX Modeler]<br /> * [https://www.ikmultimedia.com/products/tonex/index.php?p=editor TONEX Editor]<br /> * [https://www.tone.net/ ToneNET]<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Neural Amp Modeler]]<br /> * [[Capture Types]]<br /> * [[Creating a Capture]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss TONEX captures, Tone Models, hardware, and workflows on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:TONEX]]<br /> [[Category:Capture technology]]<br /> [[Category:Commercial capture systems]]</div> Tue, 01 Sep 2026 06:04:15 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:TONEX https://namforum.com/wiki/index.php?title=NAM_Playback&diff=15&oldid=0 https://namforum.com/wiki/index.php?title=NAM_Playback&diff=15&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= NAM Playback =<br /> <br /> &#039;&#039;&#039;NAM playback&#039;&#039;&#039; is the real-time execution of a trained [[Neural Amp Modeler]] model.<br /> <br /> A trained &#039;&#039;.nam&#039;&#039; file contains the neural-network model, but the file does not process audio by itself. It must be loaded into compatible software or hardware that implements the model architecture.<br /> <br /> NAM models can be played through:<br /> <br /> * Audio plugins<br /> * Standalone computer applications<br /> * Dedicated hardware<br /> * Multi-effects processors<br /> * Embedded systems<br /> * Third-party applications incorporating NAM technology<br /> <br /> Because NAM&#039;s real-time DSP implementation is open source, the same model format can be implemented by many independent developers and manufacturers.<br /> <br /> == Playback signal flow ==<br /> <br /> A basic NAM playback chain is:<br /> <br /> &#039;&#039;&#039;Instrument → audio interface → NAM player → output&#039;&#039;&#039;<br /> <br /> For a direct amplifier model, the chain may instead be:<br /> <br /> &#039;&#039;&#039;Instrument → NAM model → cabinet IR/model → output&#039;&#039;&#039;<br /> <br /> For a model that already includes a cabinet and microphone:<br /> <br /> &#039;&#039;&#039;Instrument → NAM model → output&#039;&#039;&#039;<br /> <br /> The correct configuration depends on what was included in the original capture.<br /> <br /> See:<br /> <br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> <br /> == NeuralAmpModelerCore ==<br /> <br /> &#039;&#039;&#039;NeuralAmpModelerCore&#039;&#039;&#039; is the open-source C++ DSP library maintained by the NAM project for playing NAM models in real time.<br /> <br /> It provides the underlying model-execution functionality that developers can integrate into:<br /> <br /> * Plugins<br /> * Applications<br /> * Hardware<br /> * Embedded systems<br /> <br /> This separation between the model format and playback implementation is fundamental to NAM.<br /> <br /> A &#039;&#039;.nam&#039;&#039; model is not inherently tied to one particular plugin, computer, pedal, or manufacturer.<br /> <br /> == Official NAM playback software ==<br /> <br /> The NAM project provides official software for playing NAM models.<br /> <br /> The current user-facing NAM player is &#039;&#039;&#039;Gateway&#039;&#039;&#039;.<br /> <br /> Gateway loads and plays NAM snapshot models and is available for Windows and macOS.<br /> <br /> The NAM project also maintains the underlying plugin code used to integrate NeuralAmpModelerCore into desktop audio software.<br /> <br /> Historically, the official NeuralAmpModelerPlugin has provided:<br /> <br /> * VST3 plugin support<br /> * Audio Unit support on macOS<br /> * Standalone desktop operation<br /> <br /> The exact available formats and applications may change as NAM software develops.<br /> <br /> Consult the current NAM website for downloads.<br /> <br /> == Loading a model ==<br /> <br /> The basic playback process is:<br /> <br /> # Obtain a compatible &#039;&#039;.nam&#039;&#039; model.<br /> # Open a NAM-compatible player.<br /> # Load the model.<br /> # Determine whether the model contains a cabinet.<br /> # Add cabinet processing if required.<br /> # Set appropriate input and output levels.<br /> # Play through the model.<br /> <br /> Loading the model is only one part of reproducing the original captured sound.<br /> <br /> The surrounding signal chain also matters.<br /> <br /> == Input level ==<br /> <br /> Input level is particularly important with NAM because most NAM models represent nonlinear equipment.<br /> <br /> Increasing the signal entering the model can change:<br /> <br /> * Distortion<br /> * Saturation<br /> * Compression<br /> * Attack<br /> * Sustain<br /> * Dynamic response<br /> <br /> This is different from simply changing the volume after the model.<br /> <br /> If a model was captured with calibration metadata, compatible playback systems can use that information to help reproduce the electrical level relationship of the original equipment.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Calibration-aware playback ==<br /> <br /> NAM model files can contain:<br /> <br /> &#039;&#039;&#039;input_level_dbu&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;output_level_dbu&#039;&#039;&#039;<br /> <br /> metadata.<br /> <br /> These describe the relationship between digital levels and the analog levels used during capture.<br /> <br /> A calibration-aware playback system can use this information to help determine how strongly the virtual equipment should be driven and how its output level relates to the original device.<br /> <br /> This can make a model more portable between systems with different audio-interface calibration.<br /> <br /> Calibration does not simply determine loudness.<br /> <br /> For nonlinear equipment, incorrect input calibration can change the actual modeled behavior.<br /> <br /> See [[NAM Model File Format]].<br /> <br /> == Models without calibration metadata ==<br /> <br /> Many NAM models do not contain calibration metadata.<br /> <br /> They can still be used.<br /> <br /> In that case, the player cannot automatically know the precise analog level relationship that existed during the original capture.<br /> <br /> The user may therefore need to adjust the input level by ear or according to information supplied by the model creator.<br /> <br /> This is one reason good capture documentation remains important.<br /> <br /> == Direct amplifier models ==<br /> <br /> A direct amplifier NAM normally excludes the physical speaker cabinet and microphone.<br /> <br /> When played through studio monitors, headphones, PA speakers, or another full-range system, it will normally require:<br /> <br /> * A cabinet IR<br /> * A cabinet model<br /> * Another suitable speaker simulation<br /> <br /> A typical chain is:<br /> <br /> &#039;&#039;&#039;NAM amplifier model → cabinet IR → full-range output&#039;&#039;&#039;<br /> <br /> Without cabinet processing, a direct guitar-amplifier model may sound unusually bright, harsh, or aggressive.<br /> <br /> That does not necessarily indicate a defective model.<br /> <br /> == Models containing cabinets ==<br /> <br /> A model may already contain the response of:<br /> <br /> * Amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Microphone preamp<br /> <br /> In this case, another conventional cabinet IR is normally unnecessary.<br /> <br /> Adding one may result in unintended double cabinet processing.<br /> <br /> Before adding an IR, determine what the NAM actually contains.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Playing through a physical guitar cabinet ==<br /> <br /> A NAM amplifier model can also be used with a traditional guitar cabinet.<br /> <br /> A typical chain is:<br /> <br /> &#039;&#039;&#039;Instrument → NAM amplifier model → suitable power amplifier → guitar cabinet&#039;&#039;&#039;<br /> <br /> In this configuration, a cabinet IR is often disabled because the physical guitar cabinet supplies the speaker and cabinet coloration.<br /> <br /> The power-amplification system must be suitable for the cabinet and intended operating level.<br /> <br /> == Full-range playback ==<br /> <br /> NAM models containing cabinet processing are commonly reproduced through:<br /> <br /> * Studio monitors<br /> * Headphones<br /> * PA systems<br /> * FRFR systems<br /> * Full-range powered speakers<br /> <br /> These systems are intended to reproduce the complete modeled signal rather than add the strong coloration of a traditional guitar loudspeaker.<br /> <br /> No real playback system is perfectly neutral, so the monitoring environment still influences what is heard.<br /> <br /> == Native NAM playback ==<br /> <br /> A system providing &#039;&#039;&#039;native NAM playback&#039;&#039;&#039; executes the NAM model itself using a compatible implementation of its neural-network architecture.<br /> <br /> NeuralAmpModelerCore is the reference open-source implementation used by many developers.<br /> <br /> In this type of system:<br /> <br /> &#039;&#039;&#039;.nam model → NAM-compatible DSP implementation → audio&#039;&#039;&#039;<br /> <br /> The objective is to execute the model represented by the NAM file directly.<br /> <br /> == Converted NAM playback ==<br /> <br /> Some products allow a &#039;&#039;.nam&#039;&#039; file to be imported but do not execute the NAM neural network directly.<br /> <br /> Instead, the NAM model is analyzed and converted or recaptured into another model architecture that the device can execute.<br /> <br /> The process is conceptually:<br /> <br /> &#039;&#039;&#039;.nam model → conversion → proprietary model → audio&#039;&#039;&#039;<br /> <br /> This can be extremely useful, particularly on hardware that lacks the processing resources required for native NAM execution.<br /> <br /> However, it is technically different from running the original NAM model.<br /> <br /> The NAM project has described this distinction using an analogy to lossless and lossy audio conversion.<br /> <br /> A converted model attempts to reproduce the NAM but may not produce mathematically identical processing.<br /> <br /> == Why the distinction matters ==<br /> <br /> Both native and converted systems can provide useful access to the NAM ecosystem.<br /> <br /> The distinction becomes important when discussing:<br /> <br /> * Model accuracy<br /> * Architecture compatibility<br /> * CPU requirements<br /> * A1 versus A2 support<br /> * Troubleshooting<br /> * Comparisons between playback devices<br /> <br /> A product that can &#039;&#039;&#039;import a .nam file&#039;&#039;&#039; should not automatically be assumed to execute that NAM model natively.<br /> <br /> When this distinction matters, consult the manufacturer&#039;s technical documentation.<br /> <br /> == A1 playback ==<br /> <br /> The original standard NAM WaveNet architectures are now referred to collectively as &#039;&#039;&#039;A1&#039;&#039;&#039;.<br /> <br /> These include:<br /> <br /> * A1-standard<br /> * A1-lite<br /> * A1-feather<br /> * A1-nano<br /> <br /> Large libraries of A1 models exist, and current NAM implementations continue to support them.<br /> <br /> A1 remains an important part of the NAM ecosystem.<br /> <br /> == A2 playback ==<br /> <br /> [[NAM A2 Architecture|A2]] became NAM&#039;s new standard architecture on June 2, 2026.<br /> <br /> A2-capable playback requires an implementation updated to understand the newer architecture.<br /> <br /> The NAM project identified the following versions at the A2 release as supporting the architecture:<br /> <br /> * neural-amp-modeler v0.13.0<br /> * NeuralAmpModelerCore v0.5.2<br /> * NeuralAmpModelerPlugin v0.7.14<br /> <br /> Later versions should normally retain support.<br /> <br /> Older NAM software or hardware that predates A2 support may need an update before it can load A2 models.<br /> <br /> == Backward compatibility ==<br /> <br /> Current NAM implementations are designed to retain support for existing A1 models.<br /> <br /> Therefore:<br /> <br /> &#039;&#039;&#039;A2-capable player → A2 model: supported&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;A2-capable player → A1 model: supported&#039;&#039;&#039;<br /> <br /> An older A1-only player, however, cannot automatically be expected to understand A2.<br /> <br /> See [[NAM A2 Architecture]].<br /> <br /> == A2 slimmable playback ==<br /> <br /> A2 incorporates slimmable architecture capability.<br /> <br /> This allows an A2 model to be executed at different computational levels where the playback implementation exposes that capability.<br /> <br /> A player may therefore offer a higher-accuracy mode and a reduced-computation mode using the same A2 model.<br /> <br /> The exact user controls depend on the playback product.<br /> <br /> This can be particularly useful in hardware where DSP resources must be shared among multiple effects.<br /> <br /> == Hardware playback ==<br /> <br /> NAM has been implemented in a growing range of dedicated and multi-effects hardware.<br /> <br /> Hardware implementations can differ substantially.<br /> <br /> Possible approaches include:<br /> <br /> * Native NeuralAmpModelerCore execution<br /> * Independent compatible NAM implementation<br /> * Conversion into another neural model format<br /> <br /> Hardware may also differ in:<br /> <br /> * Supported NAM architectures<br /> * Maximum number of simultaneous models<br /> * Sample rate<br /> * Cabinet processing<br /> * Model storage<br /> * Calibration support<br /> * CPU allocation<br /> * Model-management workflow<br /> <br /> Therefore, &quot;NAM compatible&quot; does not necessarily describe one identical set of capabilities.<br /> <br /> == HeadRush example ==<br /> <br /> In August 2026, Firmware 5.1 added native A1 and A2 NAM playback to the HeadRush:<br /> <br /> * Prime<br /> * Core<br /> * Flex Prime<br /> <br /> NAM models can be transferred from a computer or accessed through TONE3000 integration.<br /> <br /> The implementation also demonstrates A2&#039;s slimmable capability through a Lite Mode that reduces processor usage.<br /> <br /> This is one example of a commercial device executing current NAM architectures directly.<br /> <br /> == Converted hardware implementations ==<br /> <br /> Other manufacturers have used conversion approaches to make NAM models accessible on processors that run different internal architectures.<br /> <br /> Examples documented by the NAM project have included products from manufacturers such as:<br /> <br /> * Hotone<br /> * Valeton<br /> * Sonicake<br /> * Sonulab<br /> <br /> In these systems, importing a NAM can involve creating a device-specific representation of the source model.<br /> <br /> This should not automatically be interpreted as inferior or undesirable.<br /> <br /> It represents a different engineering tradeoff.<br /> <br /> The important point is to describe the implementation accurately.<br /> <br /> == Software playback ==<br /> <br /> NAM can also be incorporated into third-party software.<br /> <br /> Because the real-time DSP implementation is available under an open-source license, developers can build NAM support into:<br /> <br /> * Amp suites<br /> * Effects processors<br /> * DAWs<br /> * Standalone applications<br /> * Experimental software<br /> * Research systems<br /> <br /> Some applications may combine NAM with:<br /> <br /> * Cabinet IRs<br /> * Effects<br /> * Routing<br /> * Preset management<br /> * Model libraries<br /> * Other modeling technologies<br /> <br /> The resulting user experience can therefore vary considerably even when the same NAM model is being executed.<br /> <br /> == CPU usage ==<br /> <br /> Neural-network models require computation for every audio sample processed.<br /> <br /> CPU requirements depend on factors including:<br /> <br /> * Model architecture<br /> * Architecture configuration<br /> * Sample rate<br /> * Number of simultaneous models<br /> * Host implementation<br /> * Processor architecture<br /> * Optimization<br /> * A2 operating mode where applicable<br /> <br /> A model that runs easily on a desktop computer may represent a significant workload on an embedded processor.<br /> <br /> A2 was developed partly to improve this relationship between accuracy and computational cost.<br /> <br /> == Latency ==<br /> <br /> Real-time NAM playback is part of a larger digital audio system.<br /> <br /> Total playing latency can include:<br /> <br /> * Audio-interface input buffering<br /> * A/D conversion<br /> * Host processing buffer<br /> * NAM processing<br /> * Additional plugins<br /> * D/A conversion<br /> * Digital hardware routing<br /> <br /> The NAM model itself is therefore only one component of total system latency.<br /> <br /> For comfortable live playing, configure the complete system appropriately rather than evaluating buffer settings in isolation.<br /> <br /> == Sample rate ==<br /> <br /> A NAM model can contain sample-rate information.<br /> <br /> If the sample-rate field is absent, the current NAM file specification states that implementations should generally assume 48 kHz.<br /> <br /> Playback implementations are responsible for handling the relationship between the model&#039;s expected sample rate and the host system appropriately.<br /> <br /> See [[NAM Model File Format]].<br /> <br /> == Comparing NAM players ==<br /> <br /> When comparing two NAM playback systems, use the same model and control as many variables as possible.<br /> <br /> Check:<br /> <br /> * Native versus converted playback<br /> * A1/A2 architecture<br /> * Input calibration<br /> * Input level<br /> * Output level<br /> * Cabinet processing<br /> * IR<br /> * Sample rate<br /> * Additional EQ<br /> * Effects<br /> * Monitoring system<br /> <br /> Otherwise, an apparent difference between NAM implementations may actually come from the surrounding signal chain.<br /> <br /> == Model libraries ==<br /> <br /> The official NAM user page currently directs users to &#039;&#039;&#039;TONE3000&#039;&#039;&#039; for sharing and discovering models.<br /> <br /> Models can also be distributed independently because a &#039;&#039;.nam&#039;&#039; file is portable.<br /> <br /> A model downloaded from any source should be accompanied by enough information to determine:<br /> <br /> * What equipment it represents<br /> * Whether a cabinet is included<br /> * Model architecture<br /> * Relevant equipment settings<br /> * Calibration where available<br /> <br /> The playback system cannot infer every aspect of the original physical rig merely from the sound of the model.<br /> <br /> == Troubleshooting playback ==<br /> <br /> If a NAM model will not load, check:<br /> <br /> # Is the file valid?<br /> # Is it actually a .nam model?<br /> # Is it A1 or A2?<br /> # Does the player support that architecture?<br /> # Is the player or firmware current?<br /> # Does the device use native NAM or require conversion?<br /> <br /> If the model loads but sounds wrong, check:<br /> <br /> # Input level<br /> # Calibration<br /> # Cabinet/IR configuration<br /> # Output level<br /> # Additional processing<br /> # Playback system<br /> <br /> Do not assume immediately that the model itself is defective.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == The importance of portability ==<br /> <br /> One of NAM&#039;s defining characteristics is that model creation and model playback are separate.<br /> <br /> A capture can be:<br /> <br /> * Recorded by one person<br /> * Trained using one service<br /> * Shared through another service<br /> * Downloaded by another user<br /> * Played in software<br /> * Played in compatible hardware<br /> <br /> No single playback product is required for the model to exist.<br /> <br /> This separation between the &#039;&#039;&#039;model&#039;&#039;&#039; and the &#039;&#039;&#039;player&#039;&#039;&#039; is one of the principal reasons NAM can function as an ecosystem rather than merely as a feature of one manufacturer&#039;s product.<br /> <br /> == Official resources ==<br /> <br /> * [https://www.neuralampmodeler.com/users NAM user and playback resources]<br /> * [https://www.neuralampmodeler.com/the-code NAM open-source repositories]<br /> * [https://www.neuralampmodeler.com/post/a2-is-released A2 release announcement]<br /> * [https://www.neuralampmodeler.com/ Neural Amp Modeler official website]<br /> <br /> == See also ==<br /> <br /> * [[Neural Amp Modeler]]<br /> * [[NAM Model File Format]]<br /> * [[Training a NAM Model]]<br /> * [[NAM A2 Architecture]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss NAM players, hardware, and model compatibility on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Neural Amp Modeler]]<br /> [[Category:NAM playback]]<br /> [[Category:Capture technology]]</div> Tue, 01 Sep 2026 06:03:27 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:NAM_Playback https://namforum.com/wiki/index.php?title=NAM_A2_Architecture&diff=14&oldid=0 https://namforum.com/wiki/index.php?title=NAM_A2_Architecture&diff=14&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= NAM A2 Architecture =<br /> <br /> &#039;&#039;&#039;A2&#039;&#039;&#039; is the second-generation standard neural-network architecture for [[Neural Amp Modeler]] snapshot models.<br /> <br /> A2 was officially released on &#039;&#039;&#039;June 2, 2026&#039;&#039;&#039; and became NAM&#039;s new standard recipe for modeling guitar and bass amplifiers and pedals.<br /> <br /> It supersedes the original standard NAM WaveNet family, now collectively referred to as &#039;&#039;&#039;A1&#039;&#039;&#039;, while preserving support for existing A1 models.<br /> <br /> A2 was developed with three major goals:<br /> <br /> * Maintain or improve modeling accuracy<br /> * Reduce computational requirements<br /> * Provide a more flexible architecture for software and hardware implementations<br /> <br /> == A1 and A2 ==<br /> <br /> Before A2, NAM&#039;s commonly used WaveNet model sizes were known as:<br /> <br /> * Standard<br /> * Lite<br /> * Feather<br /> * Nano<br /> <br /> During development of A2, these earlier architectures were renamed:<br /> <br /> * A1-standard<br /> * A1-lite<br /> * A1-feather<br /> * A1-nano<br /> <br /> The term &#039;&#039;&#039;A1&#039;&#039;&#039; therefore refers retrospectively to the original standard NAM architecture family.<br /> <br /> A2 is the succeeding generation.<br /> <br /> Existing A1 models did not become obsolete when A2 was introduced.<br /> <br /> == Why A2 was developed ==<br /> <br /> NAM began as an open-source research and software project, but its models eventually appeared across a much wider range of systems.<br /> <br /> NAM playback now occurs on:<br /> <br /> * Windows computers<br /> * macOS computers<br /> * Linux computers<br /> * Single-board computers<br /> * Embedded processors<br /> * Dedicated guitar hardware<br /> * Web services<br /> * Commercial audio products<br /> <br /> These systems have very different computational capabilities.<br /> <br /> An architecture appropriate for a desktop computer may consume too much processing power in an embedded device, while an architecture optimized solely for very small hardware might unnecessarily compromise accuracy on more capable systems.<br /> <br /> A2 was developed with these different environments in mind.<br /> <br /> == Industry-informed development ==<br /> <br /> A2 was deliberately developed with input from companies and developers implementing NAM in real products.<br /> <br /> During development, participating builders measured candidate models on their target hardware and reported computational performance.<br /> <br /> This allowed architecture decisions to be evaluated across different processors rather than optimized solely on the developer&#039;s computer.<br /> <br /> This was a significant change from the development process of the original A1 architecture.<br /> <br /> == Computational efficiency ==<br /> <br /> One of the primary objectives of A2 was improved computational efficiency.<br /> <br /> The development process evaluated the relationship between:<br /> <br /> * Model accuracy<br /> * CPU usage<br /> * Model architecture<br /> * Model size<br /> * Target hardware<br /> <br /> The goal was not simply to create the smallest possible neural network.<br /> <br /> Instead, the objective was to obtain a better relationship between computational cost and modeling accuracy.<br /> <br /> This is especially important for hardware capable of running several effects or models simultaneously.<br /> <br /> == Modeling accuracy ==<br /> <br /> Reducing CPU usage is useful only if the resulting model remains sufficiently accurate.<br /> <br /> A2 development therefore evaluated both computational performance and modeling quality.<br /> <br /> NAM&#039;s stated objectives for the new generation included maintaining or improving:<br /> <br /> * Modeling accuracy<br /> * CPU efficiency<br /> * Training time<br /> <br /> The final architecture was selected through technical evaluation followed by subjective listening tests.<br /> <br /> == Evaluation using musical material ==<br /> <br /> A2 development emphasized evaluation using musical input rather than relying exclusively on synthetic test signals.<br /> <br /> Guitar direct recordings and signals processed through common pedals were used as relevant evaluation material.<br /> <br /> This reflects an important distinction:<br /> <br /> &#039;&#039;&#039;Training data and evaluation data do not necessarily serve the same purpose.&#039;&#039;&#039;<br /> <br /> A model may be trained using a carefully designed excitation signal while its practical performance is evaluated using the kind of musical material it will actually process.<br /> <br /> == Wide range of equipment ==<br /> <br /> A2 was intended to remain a general-purpose default architecture rather than requiring a different neural-network recipe for every amplifier or pedal.<br /> <br /> Development therefore considered a variety of guitar and bass equipment.<br /> <br /> The primary target remained nonlinear equipment such as:<br /> <br /> * Amplifiers<br /> * Preamps<br /> * Overdrive pedals<br /> * Distortion pedals<br /> * Related guitar and bass equipment<br /> <br /> Speaker cabinets and microphones present a different modeling problem and can often be represented efficiently using techniques such as impulse responses.<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == A2 and WaveNet ==<br /> <br /> A2 remains within the broader WaveNet-based architecture implemented by NeuralAmpModelerCore.<br /> <br /> Development of A2 required extending the capabilities of the NAM real-time DSP core so that new architecture configurations could be represented efficiently.<br /> <br /> The final A2 recipe uses capabilities introduced during this development process rather than merely changing the number of channels in an A1 model.<br /> <br /> Consequently, an older NAM player cannot necessarily run A2 simply because it can load A1 WaveNet models.<br /> <br /> == NeuralAmpModelerCore ==<br /> <br /> &#039;&#039;&#039;NeuralAmpModelerCore&#039;&#039;&#039; is the open-source real-time DSP library used to execute NAM models.<br /> <br /> A2 development required new capabilities in the core library.<br /> <br /> At release, the NAM project specified the following minimum relevant versions for A2 support:<br /> <br /> * &#039;&#039;&#039;neural-amp-modeler v0.13.0&#039;&#039;&#039;<br /> * &#039;&#039;&#039;NeuralAmpModelerCore v0.5.2&#039;&#039;&#039;<br /> * &#039;&#039;&#039;NeuralAmpModelerPlugin v0.7.14&#039;&#039;&#039;<br /> <br /> Developers implementing NAM playback should consult the current project documentation because these versions will continue to advance.<br /> <br /> == A2 is not forward-compatible with old players ==<br /> <br /> An important distinction exists between &#039;&#039;&#039;backward compatibility&#039;&#039;&#039; and &#039;&#039;&#039;forward compatibility&#039;&#039;&#039;.<br /> <br /> A current A2-capable NAM player can continue to support older A1 models.<br /> <br /> However, a player built before A2 support was implemented cannot be expected to run A2 models.<br /> <br /> In other words:<br /> <br /> &#039;&#039;&#039;New player → A1 model: supported&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;New player → A2 model: supported&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Old A1-only player → A2 model: not necessarily supported&#039;&#039;&#039;<br /> <br /> An A1-capable product generally requires a software or firmware update to add A2 support.<br /> <br /> The NAM project stated during A2 development that there is nothing inherent in A2&#039;s CPU requirements that prevents existing NAM hardware from supporting it, provided the manufacturer implements the required software changes.<br /> <br /> == Existing A1 models remain valid ==<br /> <br /> A2 does not invalidate existing NAM collections.<br /> <br /> A1 models remain usable in compatible current NAM software and hardware.<br /> <br /> This is particularly important because large libraries of A1 captures were created before A2 existed.<br /> <br /> The introduction of A2 should therefore be understood as an expansion of the NAM ecosystem rather than a replacement that makes older models unusable.<br /> <br /> == Training A2 models ==<br /> <br /> At the time of A2&#039;s release, A2 training was available through:<br /> <br /> * The official NAM Google Colab workflow<br /> * The local NAM GUI trainer<br /> * TONE3000<br /> <br /> Current simplified NAM training workflows use A2 as the standard architecture for newly trained snapshot models.<br /> <br /> See [[Training a NAM Model]].<br /> <br /> == Retraining existing captures ==<br /> <br /> A model creator who preserved the original training input and recorded output can potentially train a new A2 model from the same capture data.<br /> <br /> The physical amplifier or pedal does not necessarily need to be captured again.<br /> <br /> This demonstrates the value of preserving:<br /> <br /> * Original NAM training signal<br /> * Original returned recording<br /> * Calibration information<br /> * Equipment settings<br /> * Capture notes<br /> <br /> A capture recording is the measurement of the equipment.<br /> <br /> The neural-network model is one interpretation of that measurement.<br /> <br /> As training technology improves, preserved capture data may therefore remain useful.<br /> <br /> == A1 and A2 are separate models ==<br /> <br /> Retraining an A1 capture as A2 creates a new model.<br /> <br /> It does not transform the existing A1 model internally into A2.<br /> <br /> The A1 and A2 files represent independently trained neural networks, even when both were trained from the same recorded capture data.<br /> <br /> They may therefore exhibit small differences.<br /> <br /> For meaningful comparison, both should be evaluated using the same source material and playback chain.<br /> <br /> == Slimmable NAM research ==<br /> <br /> A2 development followed research into &#039;&#039;&#039;Slimmable NAM&#039;&#039;&#039;, a method for creating neural amp models whose computational cost can be adjusted at runtime.<br /> <br /> The concept addresses a practical problem:<br /> <br /> Different playback devices have different amounts of processing power available.<br /> <br /> A desktop computer may be able to run a large neural network easily, while an embedded processor may need a smaller computational workload.<br /> <br /> Slimmable modeling explores the possibility of allowing one trained model architecture to operate at different computational sizes.<br /> <br /> This research helped inform the goals and development direction of A2.<br /> <br /> A2 should not, however, simply be described as another name for Slimmable NAM. A2 is the standard architecture resulting from the broader architecture-development project.<br /> <br /> == Hardware support ==<br /> <br /> Because NAM is open source, A2 support can be implemented by third-party hardware manufacturers.<br /> <br /> Support depends on the firmware and NAM implementation in the particular device.<br /> <br /> A product advertised as supporting &#039;&#039;.nam&#039;&#039; files should therefore not automatically be assumed to support every NAM architecture.<br /> <br /> When evaluating hardware, determine whether it supports:<br /> <br /> * A1<br /> * A2<br /> * Both<br /> <br /> and verify the firmware version where appropriate.<br /> <br /> == Example: HeadRush ==<br /> <br /> In August 2026, HeadRush released Firmware 5.1 for the Prime, Core, and Flex Prime processors with native playback support for both NAM A2 and A1 models.<br /> <br /> This provides an example of the distinction between file-format recognition and architecture support: hardware already capable of running NAM-related workloads still required updated firmware implementing the newer architecture.<br /> <br /> Other NAM-compatible products may have different support status.<br /> <br /> Always consult current documentation for the specific product.<br /> <br /> == Identifying an A2 model ==<br /> <br /> The architecture and configuration information required to reconstruct a NAM model is stored in the &#039;&#039;.nam&#039;&#039; file.<br /> <br /> Compatible software can therefore determine the model architecture from the file itself.<br /> <br /> See [[NAM Model File Format]].<br /> <br /> A filename does not need to contain &quot;A2&quot; for the model to be an A2 model, although creators may include architecture information in filenames or descriptions for convenience.<br /> <br /> == Why architecture matters to users ==<br /> <br /> For many musicians, the internal neural-network architecture is not something that must be understood in detail.<br /> <br /> However, architecture matters when:<br /> <br /> * A model will not load<br /> * A hardware device supports A1 but not A2<br /> * CPU usage is important<br /> * Models are being trained<br /> * A1 and A2 versions are being compared<br /> * Software or firmware is being updated<br /> * Models are being distributed to other users<br /> <br /> When troubleshooting a model that refuses to load, checking architecture compatibility should be one of the first steps.<br /> <br /> == Why architecture matters to developers ==<br /> <br /> For developers, architecture support determines how the model must be reconstructed and executed.<br /> <br /> A NAM host must understand the architecture and configuration contained in the model file and implement the required processing correctly.<br /> <br /> A2 development was conducted partly in cooperation with hardware implementers specifically to make the new standard practical across a wide range of devices.<br /> <br /> The authoritative implementation remains the open-source NAM code.<br /> <br /> == A2 development timeline ==<br /> <br /> Major public milestones included:<br /> <br /> &#039;&#039;&#039;November 2025&#039;&#039;&#039;<br /> <br /> Slimmable NAM research was presented and the concept of a future A2 standard was discussed.<br /> <br /> &#039;&#039;&#039;January 17, 2026&#039;&#039;&#039;<br /> <br /> The formal Architecture A2 development project was announced.<br /> <br /> &#039;&#039;&#039;January 2026&#039;&#039;&#039;<br /> <br /> New NeuralAmpModelerCore capabilities and hardware test models were released for developers.<br /> <br /> &#039;&#039;&#039;March 2026&#039;&#039;&#039;<br /> <br /> Optimization work produced early candidate A2 designs.<br /> <br /> &#039;&#039;&#039;April 2026&#039;&#039;&#039;<br /> <br /> Development progressed to final listening tests and updated NeuralAmpModelerCore releases.<br /> <br /> &#039;&#039;&#039;June 2, 2026&#039;&#039;&#039;<br /> <br /> A2 was officially released as NAM&#039;s new standard architecture.<br /> <br /> == Practical guidance ==<br /> <br /> For newly trained NAM snapshot models, A2 is the current standard.<br /> <br /> Keep older A1 models. They remain useful and supported.<br /> <br /> When distributing A2 models, remember that some older software and hardware may require an update before they can play them.<br /> <br /> When a NAM model fails to load, determine:<br /> <br /> # Whether the model is A1 or A2.<br /> # Whether the player supports that architecture.<br /> # Whether the software or firmware is current.<br /> # Whether the file itself is valid.<br /> <br /> Do not assume that failure to load an A2 model means the &#039;&#039;.nam&#039;&#039; file is corrupt.<br /> <br /> == Official resources ==<br /> <br /> * [https://www.neuralampmodeler.com/post/a2-is-released A2 release announcement]<br /> * [https://www.neuralampmodeler.com/post/architecture-a2 Architecture A2 development]<br /> * [https://www.neuralampmodeler.com/ Neural Amp Modeler official website]<br /> <br /> == See also ==<br /> <br /> * [[Neural Amp Modeler]]<br /> * [[NAM Model File Format]]<br /> * [[Training a NAM Model]]<br /> * [[NAM Playback]]<br /> * [[Capture Types]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss NAM A1, A2, training, and compatibility on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Neural Amp Modeler]]<br /> [[Category:NAM architectures]]<br /> [[Category:Capture technology]]</div> Tue, 01 Sep 2026 06:02:30 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:NAM_A2_Architecture https://namforum.com/wiki/index.php?title=Training_a_NAM_Model&diff=13&oldid=0 https://namforum.com/wiki/index.php?title=Training_a_NAM_Model&diff=13&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Training a NAM Model =<br /> <br /> &#039;&#039;&#039;Training a NAM model&#039;&#039;&#039; is the process of using recorded input and output audio to create a [[Neural Amp Modeler]] model of an amplifier, pedal, preamp, or other nonlinear audio system.<br /> <br /> NAM learns the relationship between a known input signal and the signal produced after that input passes through the equipment being modeled.<br /> <br /> The basic process is:<br /> <br /> &#039;&#039;&#039;NAM input file → device under test → recorded output file → NAM trainer → .nam model&#039;&#039;&#039;<br /> <br /> Creating a good model therefore begins before training. The quality and accuracy of the recorded data are fundamental to the result.<br /> <br /> == Before training ==<br /> <br /> Before creating training data, decide exactly what the model should represent.<br /> <br /> The device under test may be:<br /> <br /> * An amplifier<br /> * A pedal<br /> * A preamp<br /> * A direct amplifier signal<br /> * An amplifier and cabinet<br /> * Multiple devices in a chain<br /> <br /> Configure the equipment and document its settings.<br /> <br /> See:<br /> <br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Capture Signal Chain]]<br /> <br /> == Obtain the official input file ==<br /> <br /> The simplified NAM training workflow provides a standardized input audio file.<br /> <br /> The current local GUI trainer includes a &#039;&#039;&#039;Download input file&#039;&#039;&#039; button that links to the input audio used for training.<br /> <br /> Use the input file intended for the version and workflow of the NAM trainer being used.<br /> <br /> Do not casually modify the training file.<br /> <br /> Changes to its:<br /> <br /> * Level<br /> * Length<br /> * Sample rate<br /> * Bit depth<br /> * Timing<br /> * Content<br /> <br /> can interfere with training or alter the relationship between the expected input and recorded output.<br /> <br /> == Reamp the input through the equipment ==<br /> <br /> The NAM input file is sent through the equipment being modeled.<br /> <br /> A typical path is:<br /> <br /> &#039;&#039;&#039;NAM input.wav → audio interface output → reamp/capture path → device under test → return path → audio interface input&#039;&#039;&#039;<br /> <br /> The resulting signal is recorded as the output file.<br /> <br /> See [[Reamping for Capture]].<br /> <br /> Depending on the equipment, the return path might be:<br /> <br /> &#039;&#039;&#039;Pedal → interface input&#039;&#039;&#039;<br /> <br /> or:<br /> <br /> &#039;&#039;&#039;Amplifier → suitable load/direct capture device → interface input&#039;&#039;&#039;<br /> <br /> or:<br /> <br /> &#039;&#039;&#039;Amplifier → cabinet → microphone → preamp/interface input&#039;&#039;&#039;<br /> <br /> The recorded output contains the behavior NAM will attempt to learn.<br /> <br /> == File format ==<br /> <br /> For the current simplified NAM GUI training workflow, the official input file is:<br /> <br /> * &#039;&#039;&#039;48 kHz&#039;&#039;&#039;<br /> * &#039;&#039;&#039;24-bit&#039;&#039;&#039;<br /> * &#039;&#039;&#039;WAVE&#039;&#039;&#039;<br /> <br /> The rendered output should match those characteristics.<br /> <br /> The output recording must also match the length of the supplied input file.<br /> <br /> These requirements should be checked against the current NAM documentation whenever the trainer or workflow changes.<br /> <br /> == Record the return cleanly ==<br /> <br /> Unless intentionally part of the device being modeled, avoid adding processing to the recorded return.<br /> <br /> Check for:<br /> <br /> * DAW plugins<br /> * EQ<br /> * Compression<br /> * Limiting<br /> * Normalization<br /> * Noise reduction<br /> * Cabinet simulation<br /> * Interface DSP<br /> * Automatic gain control<br /> * Master-bus processing<br /> <br /> Anything in the measured path can influence the model.<br /> <br /> If processing is deliberately part of the target chain, document it.<br /> <br /> == Avoid clipping ==<br /> <br /> The return recording must not contain unintended clipping.<br /> <br /> Check every stage of the capture path rather than relying only on the final DAW meter.<br /> <br /> Possible overload points include:<br /> <br /> * Interface output<br /> * Reamp device<br /> * Device under test<br /> * Load/direct-output equipment<br /> * Microphone preamp<br /> * Interface input<br /> * A/D converter<br /> * DAW signal path<br /> <br /> Intentional distortion produced by the equipment being modeled is different from unintended clipping elsewhere in the recording system.<br /> <br /> == Send level ==<br /> <br /> The NAM calibration documentation recommends providing the equipment with training examples at least as loud as the levels at which the resulting model is expected to be used.<br /> <br /> This is particularly important with nonlinear equipment.<br /> <br /> If the training signal never drives the equipment into a region that will later be demanded from the model, the trainer has no measured examples of that behavior from which to learn.<br /> <br /> This does not mean that arbitrary overload is desirable.<br /> <br /> The objective is to expose the equipment to the useful operating range that the model is expected to reproduce while maintaining a valid capture signal chain.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Calibration ==<br /> <br /> NAM supports optional calibration metadata describing the relationship between digital signal level and the analog levels used during capture.<br /> <br /> The relevant metadata fields are:<br /> <br /> &#039;&#039;&#039;input_level_dbu&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;output_level_dbu&#039;&#039;&#039;<br /> <br /> Providing these values is optional. A NAM model can operate without calibration metadata.<br /> <br /> However, calibrated models can allow compatible playback systems to reproduce the gain relationship of the original analog equipment more accurately.<br /> <br /> The official NAM calibration procedure involves measuring the analog send level and determining the return-interface level corresponding to digital full scale.<br /> <br /> See:<br /> <br /> * [[Gain Staging and Calibration]]<br /> * [[NAM Model File Format]]<br /> <br /> == Do not change hardware gain after calibration measurement ==<br /> <br /> Calibration describes the actual hardware configuration used to create the training data.<br /> <br /> If output or input gain controls are changed after recording, calibration measurements made using the new settings no longer describe the original capture.<br /> <br /> For repeatable work, document and preserve:<br /> <br /> * Interface output setting<br /> * Reamp setting<br /> * Interface input gain<br /> * Other variable gain stages<br /> <br /> until the relevant measurements are complete.<br /> <br /> == Preserve the original recording ==<br /> <br /> Keep an untouched copy of the returned training recording.<br /> <br /> This is useful if:<br /> <br /> * Training must be repeated<br /> * A new architecture becomes available<br /> * Metadata must be corrected<br /> * A problem is discovered<br /> * Another trainer is used<br /> * Training parameters are compared<br /> <br /> Do not make destructive edits to the only copy of the original capture data.<br /> <br /> == Input and output length ==<br /> <br /> The current simplified NAM trainer requires the returned output recording to match the length of the input file.<br /> <br /> Problems can arise if:<br /> <br /> * Recording starts late<br /> * Recording stops early<br /> * Extra silence is added<br /> * Material is removed<br /> * The DAW adds a tail<br /> * Export boundaries are wrong<br /> <br /> If editing is required to meet the trainer&#039;s specification, preserve the original recording separately.<br /> <br /> == Latency alignment ==<br /> <br /> A recording system introduces round-trip latency.<br /> <br /> This means the recorded response does not necessarily begin at exactly the same sample position as the original training signal.<br /> <br /> The current NAM GUI trainer automatically attempts to align the input and output recordings.<br /> <br /> The official NAM input file contains two impulses near its beginning that assist this process.<br /> <br /> During training, the GUI displays its detected alignment so the user can inspect whether the response and input appear correctly aligned.<br /> <br /> DAWs may also perform latency compensation, and in some configurations that compensation can be incorrect or excessive.<br /> <br /> Therefore, do not assume that DAW compensation automatically guarantees correct alignment.<br /> <br /> == Select the files ==<br /> <br /> In the local GUI trainer, select:<br /> <br /> * The NAM input file<br /> * The recorded output file<br /> * The location where the trained model should be saved<br /> <br /> Once valid files have been selected, the trainer enables the training process.<br /> <br /> The trainer checks the files for problems before training begins.<br /> <br /> == Batch training ==<br /> <br /> The NAM GUI supports training multiple models in a batch.<br /> <br /> Multiple reamped output files can be selected while using the corresponding input training file.<br /> <br /> This can be useful when several captures have been made from:<br /> <br /> * Different amplifiers<br /> * Different gain settings<br /> * Different pedal settings<br /> * Different channels<br /> * Different signal chains<br /> <br /> Batch training can reduce repetitive setup when processing a larger capture session.<br /> <br /> == Training ==<br /> <br /> Start training using the trainer.<br /> <br /> The software analyzes the relationship between the known input signal and the recorded response and optimizes a neural-network model to reproduce that behavior.<br /> <br /> Training involves repeatedly adjusting the model parameters to reduce the difference between:<br /> <br /> * The actual recorded output<br /> * The output predicted by the model<br /> <br /> When training completes, the resulting model is exported as a &#039;&#039;.nam&#039;&#039; file.<br /> <br /> See [[NAM Model File Format]].<br /> <br /> == Training evaluation ==<br /> <br /> At the end of the current GUI training process, NAM displays a comparison between the model&#039;s prediction and the recorded response.<br /> <br /> This provides useful information about how closely the trained model follows the captured data.<br /> <br /> However, a successful training run is not by itself proof that the model is perceptually accurate.<br /> <br /> The model should still be evaluated by listening.<br /> <br /> == Listen to the model ==<br /> <br /> Load the resulting &#039;&#039;.nam&#039;&#039; file into a compatible NAM player.<br /> <br /> Compare it with the original equipment under controlled conditions.<br /> <br /> Where practical:<br /> <br /> # Use the same source performance.<br /> # Match playback levels.<br /> # Use equivalent cabinet processing.<br /> # Switch between original and model quickly.<br /> # Test different playing dynamics.<br /> # Test more than one instrument or source where appropriate.<br /> <br /> Listen for differences in:<br /> <br /> * Gain<br /> * Distortion<br /> * Dynamics<br /> * Attack<br /> * Compression<br /> * Low-frequency response<br /> * High-frequency response<br /> * Sustain<br /> * Noise<br /> * Transient behavior<br /> <br /> A model can achieve a technically successful training result while still revealing audible differences in practical use.<br /> <br /> == Model architecture ==<br /> <br /> NAM has supported multiple neural-network architectures.<br /> <br /> The architecture affects factors including:<br /> <br /> * Computational requirements<br /> * Model size<br /> * Training behavior<br /> * Playback compatibility<br /> * Modeling performance<br /> <br /> Current simplified NAM training workflows use the newer A2 generation for snapshot modeling.<br /> <br /> Older A1 models remain widely used.<br /> <br /> See [[NAM A2 Architecture]].<br /> <br /> == Model metadata ==<br /> <br /> When training a model for distribution, provide useful metadata where the trainer supports it.<br /> <br /> This can include:<br /> <br /> * Model name<br /> * Creator<br /> * Equipment manufacturer<br /> * Equipment model<br /> * Equipment type<br /> * Tone type<br /> * Input calibration<br /> * Output calibration<br /> <br /> Metadata makes the model substantially more useful after it has been separated from the original capture session.<br /> <br /> See [[NAM Model File Format]].<br /> <br /> == Naming models ==<br /> <br /> Use filenames and model names that remain understandable outside the original session.<br /> <br /> Useful identifiers may include:<br /> <br /> * Manufacturer<br /> * Equipment model<br /> * Channel<br /> * Gain setting<br /> * Cabinet status<br /> * Significant switches or settings<br /> <br /> Avoid relying on names such as:<br /> <br /> &#039;&#039;&#039;test1.nam&#039;&#039;&#039;<br /> <br /> or:<br /> <br /> &#039;&#039;&#039;good-one-final2.nam&#039;&#039;&#039;<br /> <br /> for models intended to become part of a permanent collection.<br /> <br /> == Retraining ==<br /> <br /> One advantage of preserving the original input and output recordings is that the same capture data may be trained again.<br /> <br /> Reasons for retraining can include:<br /> <br /> * New NAM architecture<br /> * Updated trainer<br /> * Different training settings<br /> * Improved training software<br /> * Comparison of architectures<br /> * Corrected metadata<br /> <br /> The physical equipment does not necessarily have to be recaptured if the original measurement data remains suitable for the new training method.<br /> <br /> == Troubleshooting training ==<br /> <br /> If the trainer rejects the files or the resulting model behaves incorrectly, check:<br /> <br /> * Correct input file<br /> * Correct output file<br /> * Sample rate<br /> * Bit depth<br /> * File length<br /> * Clipping<br /> * Latency alignment<br /> * Capture routing<br /> * Calibration<br /> * Unwanted processing<br /> * Software version<br /> <br /> Do not randomly modify the recording simply to make training complete.<br /> <br /> Identify which requirement is not being met.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Reproducibility ==<br /> <br /> For serious capture work, preserve enough information to reproduce the model.<br /> <br /> Useful records include:<br /> <br /> * Original NAM input file<br /> * Original recorded output<br /> * Device settings<br /> * Complete capture signal chain<br /> * Interface<br /> * Reamp device<br /> * Calibration<br /> * Trainer version<br /> * Model architecture<br /> * Metadata<br /> * Final .nam file<br /> <br /> The training file and resulting model are only part of the experiment.<br /> <br /> Good documentation allows the result to be understood later.<br /> <br /> == Official resources ==<br /> <br /> * [https://neural-amp-modeler.readthedocs.io/en/latest/tutorials/gui.html NAM local GUI training documentation]<br /> * [https://neural-amp-modeler.readthedocs.io/en/stable/tutorials/calibration.html NAM calibration documentation]<br /> * [https://www.neuralampmodeler.com/ Neural Amp Modeler official website]<br /> <br /> == See also ==<br /> <br /> * [[Neural Amp Modeler]]<br /> * [[NAM Model File Format]]<br /> * [[NAM A2 Architecture]]<br /> * [[NAM Playback]]<br /> * [[Creating a Capture]]<br /> * [[Capture Signal Chain]]<br /> * [[Reamping for Capture]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss NAM training and capture workflows on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Neural Amp Modeler]]<br /> [[Category:Model training]]<br /> [[Category:Capture techniques]]</div> Tue, 01 Sep 2026 06:01:34 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Training_a_NAM_Model https://namforum.com/wiki/index.php?title=NAM_Model_File_Format&diff=12&oldid=0 https://namforum.com/wiki/index.php?title=NAM_Model_File_Format&diff=12&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= NAM Model File Format =<br /> <br /> The &#039;&#039;&#039;NAM model file format&#039;&#039;&#039; is the file format used to store trained [[Neural Amp Modeler]] models.<br /> <br /> NAM model files normally use the:<br /> <br /> &#039;&#039;&#039;.nam&#039;&#039;&#039;<br /> <br /> file extension.<br /> <br /> A &#039;&#039;.nam&#039;&#039; file contains the information required by compatible NAM software or hardware to reconstruct and run the trained neural-network model. It can also contain metadata describing the model, the equipment that was modeled, training information, and capture calibration.<br /> <br /> == File structure ==<br /> <br /> The current NAM file format can be parsed as &#039;&#039;&#039;JSON&#039;&#039;&#039;.<br /> <br /> At the highest level, a NAM file contains a dictionary of fields describing the model.<br /> <br /> Core fields include:<br /> <br /> * &#039;&#039;&#039;version&#039;&#039;&#039;<br /> * &#039;&#039;&#039;architecture&#039;&#039;&#039;<br /> * &#039;&#039;&#039;config&#039;&#039;&#039;<br /> * &#039;&#039;&#039;weights&#039;&#039;&#039;<br /> <br /> Optional fields include:<br /> <br /> * &#039;&#039;&#039;sample_rate&#039;&#039;&#039;<br /> * &#039;&#039;&#039;metadata&#039;&#039;&#039;<br /> <br /> The exact contents of some fields depend on the architecture of the model.<br /> <br /> == Version ==<br /> <br /> The &#039;&#039;&#039;version&#039;&#039;&#039; field identifies the version of the NAM file specification.<br /> <br /> The version follows semantic versioning.<br /> <br /> This allows playback implementations to determine how the contents of a NAM file should be interpreted as the format evolves.<br /> <br /> == Architecture ==<br /> <br /> The &#039;&#039;&#039;architecture&#039;&#039;&#039; field identifies the high-level neural-network architecture used by the model.<br /> <br /> Historically, common values have included architectures such as:<br /> <br /> * WaveNet<br /> * LSTM<br /> <br /> Other architectures supported by the NAM code can also be represented.<br /> <br /> The architecture is important because the configuration and weights must be interpreted according to the model architecture that created them.<br /> <br /> See [[NAM A2 Architecture]].<br /> <br /> == Config ==<br /> <br /> The &#039;&#039;&#039;config&#039;&#039;&#039; field contains architecture-specific configuration information.<br /> <br /> Depending on the architecture, this can describe characteristics such as:<br /> <br /> * Layer configuration<br /> * Channel configuration<br /> * Receptive-field structure<br /> * Other parameters required to reconstruct the model<br /> <br /> The meaning of the configuration data therefore depends on the architecture specified in the file.<br /> <br /> == Weights ==<br /> <br /> The &#039;&#039;&#039;weights&#039;&#039;&#039; field contains the trained numerical parameters of the neural network.<br /> <br /> These values are the result of the training process.<br /> <br /> During playback, the NAM implementation reconstructs the appropriate model architecture and loads these trained parameters into it.<br /> <br /> The relationship between individual weights and the model depends on the architecture.<br /> <br /> == Sample rate ==<br /> <br /> A NAM file can contain a &#039;&#039;&#039;sample_rate&#039;&#039;&#039; field specifying the sample rate expected by the model.<br /> <br /> The value is expressed in samples per second (Hz).<br /> <br /> If the field is absent, the NAM specification states that implementations should generally assume &#039;&#039;&#039;48 kHz&#039;&#039;&#039;.<br /> <br /> Playback software or hardware must account appropriately for the sample rate expected by the model.<br /> <br /> == Metadata ==<br /> <br /> NAM files can contain a &#039;&#039;&#039;metadata&#039;&#039;&#039; dictionary with descriptive information about the model.<br /> <br /> Metadata is not merely cosmetic. Good metadata can make large model libraries substantially easier to identify, organize, search, and use correctly.<br /> <br /> Supported metadata can include information about:<br /> <br /> * Model name<br /> * Creator<br /> * Modeled equipment<br /> * Tone<br /> * Training<br /> * Calibration<br /> * Export date<br /> <br /> Not every NAM file contains all available metadata.<br /> <br /> == Model name ==<br /> <br /> The metadata field:<br /> <br /> &#039;&#039;&#039;name&#039;&#039;&#039;<br /> <br /> can contain the name of the model.<br /> <br /> Compatible playback software may use this as the model&#039;s displayed name rather than relying entirely on the filename.<br /> <br /> == Model creator ==<br /> <br /> The:<br /> <br /> &#039;&#039;&#039;modeled_by&#039;&#039;&#039;<br /> <br /> field identifies the person or organization that created the model.<br /> <br /> This can be useful when models are redistributed or downloaded independently of the collection in which they were originally published.<br /> <br /> == Equipment metadata ==<br /> <br /> NAM supports several fields describing the equipment represented by the model.<br /> <br /> These include:<br /> <br /> &#039;&#039;&#039;gear_make&#039;&#039;&#039;<br /> <br /> The manufacturer or make of the equipment.<br /> <br /> &#039;&#039;&#039;gear_model&#039;&#039;&#039;<br /> <br /> The model of the equipment.<br /> <br /> &#039;&#039;&#039;gear_type&#039;&#039;&#039;<br /> <br /> The type of equipment or signal chain represented by the NAM.<br /> <br /> The current specification defines gear-type values including:<br /> <br /> * amp<br /> * pedal<br /> * pedal_amp<br /> * amp_cab<br /> * amp_pedal_cab<br /> * preamp<br /> * studio<br /> <br /> These fields help answer an important playback question:<br /> <br /> &#039;&#039;&#039;What is actually contained in this model?&#039;&#039;&#039;<br /> <br /> See [[Capture Types]].<br /> <br /> == Tone type ==<br /> <br /> The metadata field:<br /> <br /> &#039;&#039;&#039;tone_type&#039;&#039;&#039;<br /> <br /> can classify the general character of the model.<br /> <br /> The current NAM specification defines values including:<br /> <br /> * clean<br /> * overdrive<br /> * crunch<br /> * hi_gain<br /> * fuzz<br /> <br /> This provides a standardized way for model-management systems to organize captures according to their general sound.<br /> <br /> == Training metadata ==<br /> <br /> Models created with the simplified NAM trainers can contain a:<br /> <br /> &#039;&#039;&#039;training&#039;&#039;&#039;<br /> <br /> metadata dictionary.<br /> <br /> This allows information about the training process to travel with the resulting model rather than being stored only in an external description.<br /> <br /> The exact training metadata may evolve as NAM development continues.<br /> <br /> == Calibration metadata ==<br /> <br /> NAM supports two particularly important calibration fields:<br /> <br /> &#039;&#039;&#039;input_level_dbu&#039;&#039;&#039;<br /> <br /> and:<br /> <br /> &#039;&#039;&#039;output_level_dbu&#039;&#039;&#039;<br /> <br /> These allow the model file to describe the analog level relationships used during capture.<br /> <br /> === input_level_dbu ===<br /> <br /> &#039;&#039;&#039;input_level_dbu&#039;&#039;&#039; represents the analog level, expressed in dBu, presented to the modeled equipment when a 1 kHz digital sine wave has a peak level of 0 dBFS.<br /> <br /> This provides a reference between the digital signal and the electrical input level used during the capture.<br /> <br /> === output_level_dbu ===<br /> <br /> &#039;&#039;&#039;output_level_dbu&#039;&#039;&#039; represents the analog level, expressed in dBu, of a 1 kHz sine wave that produces a 0 dBFS peak when fed into the recording interface used for the output of the modeled equipment.<br /> <br /> Together, these values can allow compatible systems to preserve meaningful gain relationships between capture and playback.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Why calibration metadata matters ==<br /> <br /> Consider two people using different audio interfaces.<br /> <br /> The same digital signal might produce different analog voltages from the two interfaces.<br /> <br /> Without calibration information, both users can load the same NAM model but drive it at different effective analog-equivalent levels.<br /> <br /> With appropriate calibration metadata and compatible playback equipment, the intended gain relationship can be reproduced more accurately.<br /> <br /> This is particularly important with nonlinear equipment because changing the effective input level can change:<br /> <br /> * Distortion<br /> * Saturation<br /> * Compression<br /> * Attack<br /> * Dynamic response<br /> <br /> Calibration therefore affects behavior, not merely volume.<br /> <br /> == Metadata is optional ==<br /> <br /> A NAM model does not need complete metadata in order to function.<br /> <br /> Many existing NAM files contain little descriptive information beyond the data required to run the model.<br /> <br /> However, models intended for public distribution are considerably more useful when they include or are accompanied by meaningful documentation.<br /> <br /> Useful information includes:<br /> <br /> * Equipment modeled<br /> * Equipment settings<br /> * Cabinet status<br /> * Capture method<br /> * Creator<br /> * Calibration<br /> * Model architecture<br /> * Training information<br /> <br /> A filename alone is rarely sufficient documentation for a long-lived public model library.<br /> <br /> == Sequential models ==<br /> <br /> The NAM specification also supports &#039;&#039;&#039;Sequential&#039;&#039;&#039; models.<br /> <br /> A Sequential model combines complete child NAM models in processing order.<br /> <br /> Each child model contains its own:<br /> <br /> * Architecture<br /> * Configuration<br /> * Weights<br /> <br /> The wrapper itself contains the ordered collection of child models.<br /> <br /> This makes it possible for the NAM format to describe a processing chain composed of multiple NAM-compatible model elements rather than only a single model.<br /> <br /> The sample rates of the top-level Sequential model and its child models must agree.<br /> <br /> == Portability ==<br /> <br /> One of the important characteristics of the NAM ecosystem is that the model file is separate from the system that created it.<br /> <br /> A &#039;&#039;.nam&#039;&#039; file can potentially be:<br /> <br /> * Trained by one system<br /> * Stored by another service<br /> * Distributed through a model library<br /> * Loaded into multiple software players<br /> * Loaded into compatible hardware<br /> <br /> The model is therefore not inherently tied to the computer or service that performed the training.<br /> <br /> Actual compatibility still depends on whether the playback implementation supports the architecture and file-format version used by the model.<br /> <br /> == Forward compatibility ==<br /> <br /> NAM continues to develop.<br /> <br /> New architectures, metadata fields, and file-format capabilities can be introduced over time.<br /> <br /> A NAM player that predates a particular architecture may be able to recognize the file but still be unable to execute the model.<br /> <br /> For this reason, when a model fails to load, check:<br /> <br /> * Model architecture<br /> * NAM file version<br /> * Playback software version<br /> * Hardware firmware version<br /> <br /> before assuming the file itself is corrupt.<br /> <br /> == Editing NAM files ==<br /> <br /> Because the current NAM format is JSON-parsable, portions of a NAM file can technically be inspected using tools capable of reading JSON.<br /> <br /> However, users should not casually modify model configuration or weight data.<br /> <br /> Changing architecture-specific data without understanding its meaning can make the model invalid or alter its behavior.<br /> <br /> Metadata is a more appropriate area for tools specifically designed to manage or update descriptive information.<br /> <br /> Always preserve an original copy before manually modifying a model file.<br /> <br /> == File extension and filenames ==<br /> <br /> The &#039;&#039;.nam&#039;&#039; extension identifies the file as a Neural Amp Modeler model.<br /> <br /> The filename itself is not a reliable description of what the model contains.<br /> <br /> For example, a file named after an amplifier could represent:<br /> <br /> * Amplifier only<br /> * Amplifier and cabinet<br /> * Pedal and amplifier<br /> * Amplifier, cabinet, and microphone<br /> * A larger signal chain<br /> <br /> Consult metadata and creator documentation whenever possible.<br /> <br /> See [[Capture Types]].<br /> <br /> == Official specification ==<br /> <br /> The authoritative technical specification for the NAM model file format is maintained in the official Neural Amp Modeler documentation:<br /> <br /> https://neural-amp-modeler.readthedocs.io/en/latest/model-file.html<br /> <br /> Because NAM is under active development, developers implementing NAM file support should consult the current specification rather than relying solely on secondary documentation.<br /> <br /> == See also ==<br /> <br /> * [[Neural Amp Modeler]]<br /> * [[Training a NAM Model]]<br /> * [[NAM A2 Architecture]]<br /> * [[NAM Playback]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Capture Types]]<br /> * [[Troubleshooting Captures]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss NAM models and file compatibility on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Neural Amp Modeler]]<br /> [[Category:Model formats]]<br /> [[Category:Capture technology]]</div> Tue, 01 Sep 2026 06:00:40 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:NAM_Model_File_Format https://namforum.com/wiki/index.php?title=Neural_Amp_Modeler&diff=11&oldid=0 https://namforum.com/wiki/index.php?title=Neural_Amp_Modeler&diff=11&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Neural Amp Modeler =<br /> <br /> &#039;&#039;&#039;Neural Amp Modeler (NAM)&#039;&#039;&#039; is a free and open-source system for creating highly accurate digital models of nonlinear audio equipment using deep learning.<br /> <br /> NAM is commonly used to model guitar and bass amplifiers, pedals, preamps, and related signal chains. Unlike a conventional algorithmic model designed specifically around a particular amplifier or circuit, a NAM model is trained from measured input and output audio representing the behavior of the equipment being modeled.<br /> <br /> NAM is both a technology and an open-source ecosystem. Models can be created independently, exchanged as files, and played by software and hardware that implement compatible NAM playback.<br /> <br /> == Overview ==<br /> <br /> A typical NAM modeling workflow consists of:<br /> <br /> # Obtaining the standardized NAM training signal.<br /> # Sending that signal through the equipment to be modeled.<br /> # Recording the equipment&#039;s output.<br /> # Training a neural-network model using the original input and recorded output.<br /> # Exporting the trained model as a &#039;&#039;.nam&#039;&#039; file.<br /> # Loading that model into compatible playback software or hardware.<br /> <br /> This separation between &#039;&#039;&#039;capture, training, model file, and playback&#039;&#039;&#039; is an important characteristic of the NAM ecosystem.<br /> <br /> A model does not have to be trained by the device that eventually plays it.<br /> <br /> == Open-source architecture ==<br /> <br /> NAM is an open-source project.<br /> <br /> The project separates several major functions into different software components, including:<br /> <br /> * Training code used to create models<br /> * Core DSP used to run models in real time<br /> * Plugin and standalone software used for playback<br /> <br /> This architecture allows third-party developers to incorporate NAM playback into their own software and hardware.<br /> <br /> It also allows researchers and developers to examine, modify, and extend the underlying technology rather than treating the modeling system as a closed platform.<br /> <br /> == What NAM can model ==<br /> <br /> NAM is designed for modeling nonlinear audio systems.<br /> <br /> Common targets include:<br /> <br /> * Guitar amplifiers<br /> * Bass amplifiers<br /> * Overdrive pedals<br /> * Distortion pedals<br /> * Boosts<br /> * Preamps<br /> * Combinations of nonlinear devices<br /> * Larger signal chains<br /> <br /> The practical suitability of a particular device depends on whether its behavior can be represented effectively by the model architecture and training process.<br /> <br /> NAM should not be assumed to reproduce every possible type of audio processor equally well.<br /> <br /> Time-varying effects, long delays, modulation, reverberation, and other systems with behavior fundamentally different from the equipment NAM was designed around may require different modeling approaches.<br /> <br /> == Snapshot models ==<br /> <br /> Most NAM models used for amplifiers and pedals are &#039;&#039;&#039;snapshot models&#039;&#039;&#039;.<br /> <br /> A snapshot model represents the behavior of equipment at the configuration used when the training data was recorded.<br /> <br /> For example, if an amplifier is captured with a particular combination of:<br /> <br /> * Gain<br /> * Bass<br /> * Middle<br /> * Treble<br /> * Presence<br /> * Channel<br /> * Boost or voicing switches<br /> <br /> the resulting NAM represents that measured configuration.<br /> <br /> Playback controls can alter the signal around or after the model, but they should not automatically be interpreted as equivalent to changing the corresponding controls on the original amplifier.<br /> <br /> For substantially different amplifier settings, separate NAM models are commonly created.<br /> <br /> == Training data ==<br /> <br /> NAM training requires paired input and output data.<br /> <br /> The &#039;&#039;&#039;input&#039;&#039;&#039; is the known signal sent into the equipment.<br /> <br /> The &#039;&#039;&#039;output&#039;&#039;&#039; is the response recorded after that signal passes through the equipment being modeled.<br /> <br /> The trainer uses the relationship between these signals to learn a mathematical model that predicts how the equipment transforms incoming audio.<br /> <br /> For the simplified NAM training workflow, the project provides a standardized training signal.<br /> <br /> A typical signal path is:<br /> <br /> &#039;&#039;&#039;NAM training signal → audio interface → reamp/capture path → device under test → audio interface → recorded output&#039;&#039;&#039;<br /> <br /> See:<br /> <br /> * [[Capture Signal Chain]]<br /> * [[Reamping for Capture]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Training a NAM Model]]<br /> <br /> == Local training ==<br /> <br /> NAM provides tools for training models locally on a computer.<br /> <br /> The simplified local GUI trainer allows the user to select the original NAM input file and the recorded output file and then train a model.<br /> <br /> The current documented simplified workflow uses 48 kHz, 24-bit WAV data and requires the returned recording to match the length of the supplied input file.<br /> <br /> The trainer also attempts to determine and compensate for round-trip latency between the input and output recordings.<br /> <br /> More advanced command-line training tools provide greater control over the model architecture, training data, and learning configuration.<br /> <br /> See [[Training a NAM Model]].<br /> <br /> == Cloud and third-party training ==<br /> <br /> NAM models do not have to be trained on the user&#039;s own computer.<br /> <br /> The NAM project provides a Google Colab workflow for cloud-based training.<br /> <br /> Third-party services can also train compatible NAM models. This makes it possible to create NAM captures without maintaining a local machine-learning environment.<br /> <br /> The current official NAM site identifies &#039;&#039;&#039;TONE3000&#039;&#039;&#039; as a community model library and a service capable of training NAM models.<br /> <br /> == The .nam model file ==<br /> <br /> Trained NAM models are normally distributed as files using the:<br /> <br /> &#039;&#039;&#039;.nam&#039;&#039;&#039;<br /> <br /> file extension.<br /> <br /> A &#039;&#039;.nam&#039;&#039; file contains the information required by compatible playback software to reconstruct and run the trained neural-network model.<br /> <br /> The format includes information describing the model architecture, configuration, weights, sample rate, and optional metadata.<br /> <br /> See [[NAM Model File Format]].<br /> <br /> == Model metadata ==<br /> <br /> The NAM file specification supports metadata that can provide useful information about a model and its creation.<br /> <br /> Depending on the file and trainer used, metadata may include information such as:<br /> <br /> * Model name<br /> * Model creator<br /> * Equipment manufacturer<br /> * Equipment model<br /> * Equipment type<br /> * Tone type<br /> * Training information<br /> * Input calibration<br /> * Output calibration<br /> <br /> Metadata is especially valuable when models are distributed publicly because the filename alone may provide little information about how the model was created.<br /> <br /> == Calibration metadata ==<br /> <br /> NAM supports calibration metadata describing the relationship between digital signal level and the analog levels used during capture.<br /> <br /> Two important values are:<br /> <br /> * &#039;&#039;&#039;input level&#039;&#039;&#039; — the analog level corresponding to digital full scale at the input of the modeled equipment<br /> * &#039;&#039;&#039;output level&#039;&#039;&#039; — the analog return level corresponding to digital full scale in the recording system<br /> <br /> Correct calibration can help compatible playback systems reproduce the gain relationship of the original analog equipment.<br /> <br /> Calibration metadata is optional. A model without it can still operate, but the relationship between digital playback level and the original equipment may not be accurately reproduced.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == A1 architecture ==<br /> <br /> The earlier standard NAM WaveNet architectures are now collectively referred to as &#039;&#039;&#039;A1&#039;&#039;&#039;.<br /> <br /> These include architectures historically described as:<br /> <br /> * Standard<br /> * Lite<br /> * Feather<br /> * Nano<br /> <br /> The A1 terminology was introduced to distinguish the original architecture family from the newer [[NAM A2 Architecture|A2]] generation.<br /> <br /> Existing A1 models remain part of the NAM ecosystem and continue to be supported by compatible NAM implementations.<br /> <br /> == A2 architecture ==<br /> <br /> &#039;&#039;&#039;A2&#039;&#039;&#039; is the current standard NAM architecture generation.<br /> <br /> A2 was officially released on &#039;&#039;&#039;June 2, 2026&#039;&#039;&#039;.<br /> <br /> It superseded A1 as the default architecture for newly trained snapshot models while retaining compatibility with existing A1 models.<br /> <br /> Current NAM simplified training workflows, including the official Colab and local GUI trainers, can create A2 models.<br /> <br /> A2 support requires sufficiently recent NAM playback implementations.<br /> <br /> See [[NAM A2 Architecture]].<br /> <br /> == Playback ==<br /> <br /> A &#039;&#039;.nam&#039;&#039; file does not produce audio by itself.<br /> <br /> It must be loaded into software or hardware capable of running the model.<br /> <br /> NAM playback can exist in:<br /> <br /> * Audio plugins<br /> * Standalone applications<br /> * Hardware processors<br /> * Amplifiers<br /> * Embedded products<br /> * Third-party software<br /> <br /> Because NAM&#039;s real-time DSP implementation is available as open-source code, manufacturers and developers can integrate NAM playback into their own products.<br /> <br /> Not every NAM player necessarily supports every model architecture. Newer architectures such as A2 may require updated playback software or firmware.<br /> <br /> See [[NAM Playback]].<br /> <br /> == Cabinets and impulse responses ==<br /> <br /> A NAM model may or may not include a speaker cabinet and microphone.<br /> <br /> A direct amplifier NAM can represent the amplifier without a physical cabinet response. In that case, playback normally requires:<br /> <br /> * A cabinet impulse response<br /> * A cabinet model<br /> * A physical guitar cabinet<br /> <br /> A NAM captured through a physical cabinet and microphone may already contain those characteristics.<br /> <br /> Adding another conventional cabinet IR to such a model may result in unintended double cabinet processing.<br /> <br /> See:<br /> <br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> <br /> == Model sharing ==<br /> <br /> Because NAM models are files rather than objects locked to a particular capture device, they can be exchanged between users and compatible playback systems.<br /> <br /> This has resulted in large community collections of models representing amplifiers, pedals, and other equipment.<br /> <br /> The official NAM website currently directs users to &#039;&#039;&#039;TONE3000&#039;&#039;&#039; for community model sharing and discovery.<br /> <br /> When distributing a model, useful documentation includes:<br /> <br /> * Equipment modeled<br /> * Equipment settings<br /> * Whether a cabinet is included<br /> * Capture signal chain<br /> * Calibration information<br /> * Model architecture<br /> * Creator<br /> * Relevant notes about the capture<br /> <br /> A large model library becomes considerably more useful when its contents are documented well.<br /> <br /> == NAM as an ecosystem ==<br /> <br /> NAM is unusual among capture technologies because no single manufacturer controls every stage of the process.<br /> <br /> The ecosystem can include separate developers and services for:<br /> <br /> * Capture<br /> * Training<br /> * Model hosting<br /> * Playback software<br /> * Playback hardware<br /> * Model management<br /> * Research<br /> <br /> A model created using one NAM-compatible workflow can potentially be used by many different NAM-compatible playback systems.<br /> <br /> This interoperability is one of the most significant characteristics of the platform.<br /> <br /> == Limitations ==<br /> <br /> A successful NAM training run does not guarantee a perceptually perfect model.<br /> <br /> Accuracy can be affected by:<br /> <br /> * Training data<br /> * Calibration<br /> * Noise<br /> * Clipping<br /> * Latency alignment<br /> * Model architecture<br /> * Equipment behavior<br /> * Capture signal chain<br /> * Training configuration<br /> <br /> Some equipment may also exhibit behavior that is more difficult to reproduce than others.<br /> <br /> The resulting model should therefore be evaluated against the original equipment rather than judged solely by whether training completed successfully.<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Official resources ==<br /> <br /> * [https://www.neuralampmodeler.com/ Neural Amp Modeler official website]<br /> * [https://neural-amp-modeler.readthedocs.io/ NAM documentation]<br /> * [https://www.neuralampmodeler.com/the-code NAM open-source code]<br /> * [https://www.neuralampmodeler.com/users NAM user resources]<br /> <br /> == See also ==<br /> <br /> * [[NAM Model File Format]]<br /> * [[Training a NAM Model]]<br /> * [[NAM A2 Architecture]]<br /> * [[NAM Playback]]<br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss Neural Amp Modeler on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Neural Amp Modeler]]<br /> [[Category:Capture technology]]<br /> [[Category:Machine learning]]</div> Tue, 01 Sep 2026 05:59:14 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Neural_Amp_Modeler https://namforum.com/wiki/index.php?title=Troubleshooting_Captures&diff=10&oldid=0 https://namforum.com/wiki/index.php?title=Troubleshooting_Captures&diff=10&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Troubleshooting Captures =<br /> <br /> When a capture fails, sounds wrong, or behaves differently from the original equipment, the problem is not necessarily the capture technology itself.<br /> <br /> A capture system is part of a larger measurement chain involving software, audio interfaces, routing, levels, reamping, the device under test, and the return path. A fault or configuration error at any point can affect the resulting model.<br /> <br /> The most effective troubleshooting method is to &#039;&#039;&#039;work through the signal chain one stage at a time&#039;&#039;&#039;.<br /> <br /> == Start with the original equipment ==<br /> <br /> Before investigating the capture system, verify that the equipment being captured sounds and behaves normally.<br /> <br /> Check:<br /> <br /> * Instrument and signal cables<br /> * Power supplies<br /> * Pedal power<br /> * Amplifier operation<br /> * Tubes where applicable<br /> * Switches and channel settings<br /> * Cabinet and speaker connections<br /> * Load box or direct-output equipment<br /> * Microphone and microphone preamp<br /> * Noise, hum, or intermittent connections<br /> <br /> If the original equipment is not behaving correctly, the capture cannot accurately represent it.<br /> <br /> == Verify the capture signal ==<br /> <br /> Confirm that the correct test or training signal is being used.<br /> <br /> Check:<br /> <br /> * Correct file or signal<br /> * Correct sample rate<br /> * Correct channel<br /> * Correct playback level<br /> * Correct beginning and end<br /> * No normalization<br /> * No accidental editing<br /> * No plugins or processing<br /> * No fades unless specifically required<br /> * No sample-rate conversion unless required by the workflow<br /> <br /> Some capture systems depend on the exact relationship between the supplied test signal and the recorded response.<br /> <br /> Do not substitute or modify a training signal unless the platform specifically permits it.<br /> <br /> == Verify output routing ==<br /> <br /> Confirm that the test signal is leaving through the intended physical output.<br /> <br /> DAW and interface routing can become complicated, particularly when software mixers are involved.<br /> <br /> Check:<br /> <br /> # DAW track output<br /> # Bus or master routing<br /> # Interface software mixer<br /> # Physical interface output<br /> # Hardware output level<br /> # Reamp or capture-device input<br /> <br /> If possible, verify the signal at more than one point in the chain.<br /> <br /> Do not assume that meter activity inside the DAW proves the correct physical output is carrying the signal.<br /> <br /> == Verify the signal reaching the device ==<br /> <br /> Confirm that the device under test is actually receiving the expected signal.<br /> <br /> For an amplifier or pedal, listen to or otherwise monitor its response where safe and practical.<br /> <br /> If the device receives no signal, investigate:<br /> <br /> * Output selection<br /> * Muting<br /> * Interface mixer configuration<br /> * Cable<br /> * Reamp device<br /> * Reamp output level<br /> * Incorrect connector<br /> * Pedal bypass state<br /> * Amplifier input or channel<br /> <br /> If the device receives a signal but reacts very differently from normal instrument use, investigate calibration and level.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Verify the return path ==<br /> <br /> Next confirm that the output of the device reaches the intended recording input.<br /> <br /> Check:<br /> <br /> * Physical cable<br /> * Load/direct-output device<br /> * Microphone<br /> * Microphone preamp<br /> * Interface input<br /> * Input type<br /> * Input gain<br /> * Interface software routing<br /> * DAW input assignment<br /> <br /> A common mistake is recording a different interface input from the one actually connected to the capture chain.<br /> <br /> == No returned signal ==<br /> <br /> If the capture system receives no return signal, trace the path backward from the recording software.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Recording software ← interface input ← return cable ← device output&#039;&#039;&#039;<br /> <br /> Determine the last point at which a valid signal exists.<br /> <br /> Possible causes include:<br /> <br /> * Wrong recording input<br /> * Muted interface channel<br /> * Incorrect cable<br /> * Reamp or capture device connected backward<br /> * Pedal or processor bypass/routing problem<br /> * Amplifier standby<br /> * Incorrect direct-output configuration<br /> * Load-box routing<br /> * Microphone/preamp problem<br /> <br /> Avoid changing several things simultaneously. Find the failed stage.<br /> <br /> == Input signal too low ==<br /> <br /> A return signal that is too low can result in poor signal-to-noise ratio or capture-system errors.<br /> <br /> Possible causes include:<br /> <br /> * Device output set too low<br /> * Interface input gain too low<br /> * Incorrect input type<br /> * Excessive attenuation<br /> * Incorrect load/direct-output setting<br /> * Microphone preamp gain too low<br /> * Wrong interface reference level<br /> * Calibration error<br /> <br /> Do not automatically solve every low-level problem by increasing the final interface gain. Determine where the unexpected level loss occurs.<br /> <br /> == Input signal too high ==<br /> <br /> An excessively high return signal can cause clipping.<br /> <br /> Possible clipping points include:<br /> <br /> * Device output<br /> * Load/direct-output equipment<br /> * Microphone preamp<br /> * Interface analog input<br /> * A/D converter<br /> * DAW or capture software<br /> <br /> Remember:<br /> <br /> &#039;&#039;&#039;A recorded signal below 0 dBFS does not prove that an earlier analog stage has not clipped.&#039;&#039;&#039;<br /> <br /> Reduce the level at the stage that is actually overloading rather than compensating later in the chain.<br /> <br /> == Capture sounds more or less distorted than the original ==<br /> <br /> If the resulting model has substantially different gain or distortion characteristics, investigate the level presented to the device during capture.<br /> <br /> Possible causes include:<br /> <br /> * Incorrect calibration<br /> * Interface output level changed<br /> * DAW fader changed<br /> * Reamp output changed<br /> * Interface operating level changed<br /> * Different input used<br /> * Software mixer attenuation<br /> * Incorrect capture-platform calibration value<br /> <br /> For nonlinear equipment, input level affects behavior.<br /> <br /> The problem may therefore be a calibration error rather than a defect in the model itself.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Capture is much louder or quieter than the original ==<br /> <br /> A level difference does not necessarily indicate poor modeling accuracy.<br /> <br /> Investigate:<br /> <br /> * Capture return gain<br /> * Platform output calibration<br /> * Normalization<br /> * Playback level<br /> * Model output control<br /> * Interface configuration<br /> <br /> When comparing a capture with the original equipment, level-match them carefully.<br /> <br /> Small differences in loudness can strongly influence subjective judgments of tone and quality.<br /> <br /> == Capture sounds dark or muffled ==<br /> <br /> Possible causes include:<br /> <br /> * A cabinet response already exists in the capture and another cabinet IR has been added<br /> * Unwanted cabinet simulation was active during capture<br /> * Incorrect microphone position<br /> * Incorrect direct-output configuration<br /> * Sample-rate or processing problem<br /> * EQ or filtering active in the capture chain<br /> <br /> One of the first questions should be:<br /> <br /> &#039;&#039;&#039;Does this capture already contain a cabinet?&#039;&#039;&#039;<br /> <br /> See [[Cabinets and IRs]] and [[Capture Types]].<br /> <br /> == Capture sounds excessively bright or harsh ==<br /> <br /> Possible causes include:<br /> <br /> * Direct amplifier capture being played without a cabinet or IR<br /> * Cabinet simulation accidentally bypassed<br /> * Incorrect IR<br /> * Different monitoring system<br /> * Calibration error<br /> * Incorrect microphone position<br /> * Unintended EQ<br /> * Playback configuration<br /> <br /> A direct guitar-amplifier capture without cabinet processing can sound extremely bright and unnatural when reproduced through a full-range system.<br /> <br /> That does not necessarily mean the capture is defective.<br /> <br /> == Capture sounds thin or hollow ==<br /> <br /> Possible causes include:<br /> <br /> * Phase cancellation<br /> * Multiple cabinet IRs<br /> * Misaligned IRs<br /> * Multiple microphones with timing differences<br /> * Polarity inversion<br /> * Parallel signal paths<br /> * DAW monitoring combined with direct monitoring<br /> <br /> If two versions of the same signal are being heard simultaneously with a small timing difference, comb filtering can produce a thin or hollow sound.<br /> <br /> Check for unintended parallel paths.<br /> <br /> == Hum or buzz ==<br /> <br /> Hum and buzz frequently originate from grounding rather than the capture software.<br /> <br /> Possible causes include:<br /> <br /> * Ground loops<br /> * Unbalanced cable runs<br /> * Computer and amplifier connected through different ground paths<br /> * Power supplies<br /> * Poor cables<br /> * Nearby mains equipment<br /> * Pedal power<br /> * Lighting or other electrical equipment<br /> <br /> Transformer isolation in an appropriate [[Reamping for Capture|reamp device]] can sometimes eliminate an unwanted signal-ground path.<br /> <br /> A ground-lift feature designed for the audio connection may also help.<br /> <br /> &#039;&#039;&#039;Never defeat the protective mains safety earth of an amplifier or other equipment to eliminate hum.&#039;&#039;&#039;<br /> <br /> == Computer noise or high-frequency interference ==<br /> <br /> Digital equipment can introduce characteristic noise into capture systems.<br /> <br /> Possible sources include:<br /> <br /> * USB connections<br /> * Computer power supplies<br /> * Displays<br /> * Switching power supplies<br /> * Network equipment<br /> * Ground loops<br /> * Poorly shielded cables<br /> <br /> Try to determine whether the noise enters before or after the device under test.<br /> <br /> Isolation and balanced connections can be useful diagnostic tools.<br /> <br /> == Excessive hiss ==<br /> <br /> Some amplifiers and high-gain devices naturally produce substantial hiss.<br /> <br /> Compare the capture path with the original equipment before deciding that the capture system introduced the noise.<br /> <br /> Additional hiss may come from:<br /> <br /> * Excessively low signal levels<br /> * Excessive return gain<br /> * Multiple high-gain stages<br /> * Poor signal-to-noise ratio<br /> * Pedal power supplies<br /> * Microphone preamps<br /> * Interface inputs<br /> <br /> Noise gates can complicate capture measurements and should not be introduced casually merely to make the recorded response look cleaner.<br /> <br /> == Capture contains unexpected effects ==<br /> <br /> If a model appears to contain EQ, cabinet coloration, compression, or other processing that was not intended, inspect every processing stage.<br /> <br /> Check:<br /> <br /> * DAW plugins<br /> * Interface DSP<br /> * Hardware processors<br /> * Load-box cabinet simulation<br /> * Amplifier direct-output simulation<br /> * IR loaders<br /> * Channel-strip processing<br /> * Master-bus processing<br /> <br /> Anything present in the measured path may influence the resulting model.<br /> <br /> == Training or capture process fails ==<br /> <br /> If software refuses to complete the capture or training process, first check the platform&#039;s error message and documentation.<br /> <br /> Common causes can include:<br /> <br /> * Incorrect file length<br /> * Incorrect sample rate<br /> * Incorrect number of channels<br /> * Missing or corrupted training signal<br /> * Return level outside acceptable range<br /> * Clipping<br /> * Incorrect file format<br /> * Incorrect trimming<br /> * Insufficient computer resources<br /> * Unsupported software or model version<br /> <br /> Do not repeatedly modify the recorded response at random in an attempt to make the software accept it.<br /> <br /> Determine what requirement is not being met.<br /> <br /> == Timing and file-length problems ==<br /> <br /> Some training workflows require the recorded response to correspond precisely with the supplied training signal.<br /> <br /> Problems can arise from:<br /> <br /> * Recording started too late<br /> * Recording stopped too early<br /> * Extra material before the signal<br /> * Missing material at the end<br /> * Automatic latency compensation<br /> * Manual trimming<br /> * Fades<br /> * Resampling<br /> * DAW export settings<br /> <br /> Follow the platform&#039;s specific requirements for timing and file length.<br /> <br /> If manual editing is required, preserve the original recording before modifying it.<br /> <br /> == Model works in one player but not another ==<br /> <br /> Possible causes include:<br /> <br /> * Unsupported model architecture<br /> * Older playback software<br /> * File-format version differences<br /> * Platform-specific metadata<br /> * Corrupted download<br /> * Incorrect file extension<br /> * Unsupported sample-rate or processing configuration<br /> <br /> Check the model&#039;s architecture and format against the capabilities of the playback software or hardware.<br /> <br /> This can be particularly important as capture platforms introduce new model architectures.<br /> <br /> == Model sounds different on another device ==<br /> <br /> A model may sound different when moved between playback systems even when the model itself is identical.<br /> <br /> Investigate:<br /> <br /> * Input calibration<br /> * Output level<br /> * Cabinet/IR configuration<br /> * Sample rate<br /> * Playback software<br /> * Model architecture implementation<br /> * Additional EQ or processing<br /> * Physical playback system<br /> * Guitar cabinet versus full-range playback<br /> * Input impedance<br /> * Instrument level<br /> <br /> Do not assume that loading the same model file guarantees an identical complete signal chain.<br /> <br /> == Capture differs from the original only slightly ==<br /> <br /> Perfect null-level equivalence should not automatically be expected from every capture technology.<br /> <br /> A model is an approximation of the measured system.<br /> <br /> When evaluating smaller differences, use controlled comparisons:<br /> <br /> # Use the same source performance.<br /> # Level-match the original and model.<br /> # Use the same cabinet processing where applicable.<br /> # Switch between them quickly.<br /> # Avoid changing unrelated equipment.<br /> # Repeat the test with different material.<br /> <br /> Determine whether the difference is consistent, audible in practical use, and attributable to the model rather than the comparison setup.<br /> <br /> == Change one variable at a time ==<br /> <br /> This is one of the most useful troubleshooting principles.<br /> <br /> If a capture sounds wrong and you simultaneously change:<br /> <br /> * Reamp level<br /> * Interface gain<br /> * Amplifier settings<br /> * Training settings<br /> * Cabinet IR<br /> <br /> you may obtain a better result without learning what caused the original problem.<br /> <br /> Instead:<br /> <br /> # Establish a repeatable baseline.<br /> # Change one variable.<br /> # Make another capture or measurement.<br /> # Compare the result.<br /> # Document what changed.<br /> <br /> Capture technology is particularly well suited to controlled experimentation because the same test signal can be repeated precisely.<br /> <br /> == Preserve evidence ==<br /> <br /> Before deleting an unsuccessful capture, consider preserving:<br /> <br /> * Original training response<br /> * Model file<br /> * Screenshots<br /> * Calibration values<br /> * Signal-chain notes<br /> * Equipment settings<br /> * Error messages<br /> * Software versions<br /> <br /> Today&#039;s unexplained failure can become useful evidence when a pattern emerges later.<br /> <br /> == Asking for help ==<br /> <br /> When asking other users to troubleshoot a capture, provide enough information to reproduce or understand the problem.<br /> <br /> Useful information includes:<br /> <br /> * Capture platform<br /> * Software/version<br /> * Equipment being captured<br /> * Complete signal chain<br /> * Audio interface<br /> * Reamp/capture device<br /> * Calibration method<br /> * Input and output settings<br /> * Capture type<br /> * Whether a cabinet is included<br /> * Training settings where applicable<br /> * Playback system<br /> * Description of the problem<br /> <br /> Audio examples, screenshots, measurements, and links to the model may also be useful.<br /> <br /> A detailed report is much easier to troubleshoot than simply saying:<br /> <br /> &#039;&#039;&quot;My capture doesn&#039;t sound right.&quot;&#039;&#039;<br /> <br /> == Troubleshooting checklist ==<br /> <br /> When something goes wrong, work through this sequence:<br /> <br /> # Verify the original equipment.<br /> # Verify the correct test signal.<br /> # Verify output routing.<br /> # Verify the signal reaching the device.<br /> # Verify the device settings.<br /> # Verify the return path.<br /> # Check every stage for clipping.<br /> # Check signal-to-noise ratio.<br /> # Verify calibration.<br /> # Remove unintended processing.<br /> # Verify cabinet/IR configuration.<br /> # Verify recording format and timing.<br /> # Verify training or capture settings.<br /> # Compare the model with the original under controlled conditions.<br /> # Change only one variable at a time.<br /> <br /> The objective is to identify &#039;&#039;&#039;where the first unexpected difference occurs&#039;&#039;&#039;.<br /> <br /> Once that point is known, the number of possible causes usually becomes much smaller.<br /> <br /> == See also ==<br /> <br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Ask for capture troubleshooting help on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Capture techniques]]<br /> [[Category:Troubleshooting]]</div> Tue, 01 Sep 2026 05:58:03 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Troubleshooting_Captures https://namforum.com/wiki/index.php?title=Cabinets_and_IRs&diff=9&oldid=0 https://namforum.com/wiki/index.php?title=Cabinets_and_IRs&diff=9&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Cabinets and IRs =<br /> <br /> Loudspeaker cabinets are a major part of the sound of an amplified instrument.<br /> <br /> In guitar and bass systems, the cabinet, loudspeaker, microphone, and microphone position can alter the sound at least as dramatically as many changes made to the amplifier itself.<br /> <br /> Capture technology can handle this part of the signal chain in several ways. A capture may include the cabinet and microphone, or the amplifier may be captured directly and combined later with a separate &#039;&#039;&#039;impulse response (IR)&#039;&#039;&#039; or cabinet model.<br /> <br /> Understanding which approach was used is essential when configuring a capture for playback.<br /> <br /> == What is an impulse response? ==<br /> <br /> An &#039;&#039;&#039;impulse response&#039;&#039;&#039;, commonly called an &#039;&#039;&#039;IR&#039;&#039;&#039;, is a measurement of the response of a system.<br /> <br /> In guitar and bass applications, IRs are widely used to reproduce the frequency and time-domain characteristics of a loudspeaker cabinet, speaker, microphone, and microphone position.<br /> <br /> A cabinet IR commonly represents a chain such as:<br /> <br /> &#039;&#039;&#039;Loudspeaker → cabinet → microphone → microphone position → recording path&#039;&#039;&#039;<br /> <br /> The resulting IR can be applied to another signal using a process called &#039;&#039;&#039;convolution&#039;&#039;&#039;.<br /> <br /> In practical use, this allows a direct amplifier model or capture to be combined with the measured response of a cabinet and microphone.<br /> <br /> == Captures and IRs are not the same thing ==<br /> <br /> An IR and a nonlinear equipment capture solve different problems.<br /> <br /> An amplifier, overdrive pedal, or similar device changes its behavior according to signal level. Distortion, compression, clipping, saturation, and other nonlinear behavior may change dynamically with the input.<br /> <br /> A conventional cabinet IR is fundamentally a &#039;&#039;&#039;linear&#039;&#039;&#039; representation.<br /> <br /> It can reproduce characteristics such as:<br /> <br /> * Frequency response<br /> * Phase response<br /> * Resonance<br /> * Timing information<br /> * Cabinet and speaker coloration<br /> * Microphone coloration<br /> * Microphone position<br /> <br /> It does not, by itself, reproduce the complete nonlinear behavior of an amplifier or distortion device.<br /> <br /> This is why a common signal chain is:<br /> <br /> &#039;&#039;&#039;Nonlinear amplifier capture → cabinet IR&#039;&#039;&#039;<br /> <br /> The two technologies perform complementary jobs.<br /> <br /> == Direct amplifier captures ==<br /> <br /> A [[Capture Types|direct amplifier capture]] normally excludes the physical loudspeaker cabinet and microphone.<br /> <br /> A simplified capture path might be:<br /> <br /> &#039;&#039;&#039;Amplifier → suitable load/direct capture device → capture system&#039;&#039;&#039;<br /> <br /> During playback, a cabinet IR or cabinet model can then be added:<br /> <br /> &#039;&#039;&#039;Amplifier capture → cabinet IR → output&#039;&#039;&#039;<br /> <br /> This provides substantial flexibility because the same amplifier capture can be used with many different cabinet and microphone responses.<br /> <br /> For example, one amplifier capture might be auditioned through IRs representing:<br /> <br /> * Different speaker models<br /> * Different cabinet sizes<br /> * Open-back and closed-back cabinets<br /> * Different microphones<br /> * Different microphone positions<br /> <br /> The amplifier does not need to be captured again simply to change the cabinet response.<br /> <br /> == Captures that already contain a cabinet ==<br /> <br /> A capture made using a physical cabinet and microphone may already contain the complete cabinet/microphone response.<br /> <br /> A typical path is:<br /> <br /> &#039;&#039;&#039;Amplifier → loudspeaker cabinet → microphone → preamp/interface → capture system&#039;&#039;&#039;<br /> <br /> The resulting capture may therefore contain characteristics of:<br /> <br /> * Amplifier<br /> * Speaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Microphone preamp<br /> * Other equipment in the return path<br /> <br /> Such a capture normally does &#039;&#039;&#039;not&#039;&#039;&#039; require an additional conventional cabinet IR.<br /> <br /> Adding another cabinet IR can result in two cabinet responses being applied in series.<br /> <br /> This is commonly described as &#039;&#039;&#039;double cabbing&#039;&#039;&#039; or a &#039;&#039;&#039;double cabinet&#039;&#039;&#039; configuration.<br /> <br /> The result is not necessarily unusable as a creative effect, but it is normally not an accurate reproduction of the original captured signal chain.<br /> <br /> == How cabinet IRs are created ==<br /> <br /> An IR is created by measuring the response of the system to a known signal.<br /> <br /> Several measurement techniques are possible. A common method uses a swept sine signal.<br /> <br /> A simplified measurement chain is:<br /> <br /> &#039;&#039;&#039;Test signal → power amplifier → loudspeaker cabinet → microphone → recording system&#039;&#039;&#039;<br /> <br /> The recorded response is mathematically processed to derive the impulse response.<br /> <br /> The resulting IR is usually stored as an audio file, commonly in WAV format.<br /> <br /> Once created, the IR can be loaded into compatible hardware or software.<br /> <br /> == The microphone is part of the IR ==<br /> <br /> A cabinet IR should not be thought of as merely a digital copy of a speaker.<br /> <br /> If a microphone was used to create the IR, the microphone is part of the measurement.<br /> <br /> Its contribution includes:<br /> <br /> * Frequency response<br /> * Polar response<br /> * Proximity effect<br /> * Position relative to the speaker<br /> * Angle<br /> * Distance<br /> <br /> Consequently, an IR labeled with the same cabinet and speaker can sound dramatically different when created with another microphone or microphone position.<br /> <br /> == Microphone position ==<br /> <br /> Microphone position has a major influence on cabinet sound.<br /> <br /> Moving a microphone across the face of a guitar speaker can substantially change the tonal balance.<br /> <br /> Typical variables include:<br /> <br /> * Center versus edge of the cone<br /> * Distance from the speaker<br /> * On-axis versus off-axis placement<br /> * Position relative to individual speakers in a multi-speaker cabinet<br /> <br /> For this reason, commercial and community IR libraries frequently contain many variations of the same cabinet.<br /> <br /> They are not necessarily redundant. Each may represent a genuinely different microphone configuration.<br /> <br /> == Multiple microphones ==<br /> <br /> Cabinets are often recorded using more than one microphone.<br /> <br /> An IR library may therefore contain:<br /> <br /> * Individual microphone IRs<br /> * Pre-mixed microphone combinations<br /> * Multiple positions of each microphone<br /> * Different microphone distances<br /> * Different blend ratios<br /> <br /> A mixed IR contains the characteristics of the selected microphone combination as it existed when the IR was created.<br /> <br /> If greater control is desired, separate IRs can sometimes be loaded and mixed during playback instead.<br /> <br /> == IR length ==<br /> <br /> Impulse responses can be stored at different lengths.<br /> <br /> Longer IRs can preserve more low-frequency and time-domain information, but they also require more processing and storage.<br /> <br /> Some hardware devices impose maximum IR lengths or automatically truncate or resample imported IRs.<br /> <br /> For guitar cabinet applications, the practical effect of IR length depends on the equipment, software, sample rate, and the material represented by the IR.<br /> <br /> Longer is therefore not automatically better in every application.<br /> <br /> == Sample rate ==<br /> <br /> IR files are digital audio data and have a sample rate.<br /> <br /> Playback software or hardware may:<br /> <br /> * Require a particular sample rate<br /> * Resample the IR during import<br /> * Resample it during playback<br /> * Support several rates<br /> <br /> Follow the requirements of the device or software loading the IR.<br /> <br /> An IR should not be assumed to operate differently merely because its file sample rate differs if the playback system performs appropriate conversion.<br /> <br /> == Minimum-phase and phase information ==<br /> <br /> IR libraries may offer files described as &#039;&#039;&#039;minimum phase&#039;&#039;&#039;, &#039;&#039;&#039;minimum phase transformed (MPT)&#039;&#039;&#039;, or similar terminology.<br /> <br /> Minimum-phase processing changes the phase relationship of the response while retaining its general magnitude response.<br /> <br /> This can make IRs easier to combine in some circumstances because timing differences between independently measured files are reduced.<br /> <br /> Other IRs preserve more of the original measured timing and phase information.<br /> <br /> Neither form is universally superior. The appropriate choice depends on how the IR will be used.<br /> <br /> == Combining IRs ==<br /> <br /> Two or more IRs can be blended to create a composite cabinet sound.<br /> <br /> This is commonly used to simulate multi-microphone recording.<br /> <br /> However, phase and timing relationships matter.<br /> <br /> Two individually good IRs may produce unexpected cancellations when combined.<br /> <br /> Possible results include:<br /> <br /> * Reduced low frequencies<br /> * Comb filtering<br /> * Hollow or thin sound<br /> * Changes in attack<br /> * Unexpected frequency peaks or dips<br /> <br /> IR mixing software may provide alignment, delay, phase, or polarity controls to manage these interactions.<br /> <br /> == IRs and physical guitar cabinets ==<br /> <br /> A cabinet IR is normally intended for playback through a relatively full-range system such as:<br /> <br /> * Studio monitors<br /> * Headphones<br /> * PA systems<br /> * FRFR speakers<br /> * Suitable full-range power amplifiers and speakers<br /> <br /> When a model is played through a traditional physical guitar cabinet, an additional cabinet IR may not be desirable because the physical cabinet already provides strong speaker and cabinet coloration.<br /> <br /> A common arrangement is therefore:<br /> <br /> &#039;&#039;&#039;Direct amplifier capture → power amplifier → physical guitar cabinet&#039;&#039;&#039;<br /> <br /> rather than:<br /> <br /> &#039;&#039;&#039;Direct amplifier capture → cabinet IR → power amplifier → physical guitar cabinet&#039;&#039;&#039;<br /> <br /> There are exceptions and creative uses, but understanding the signal chain prevents accidental duplication.<br /> <br /> == FRFR playback ==<br /> <br /> &#039;&#039;&#039;FRFR&#039;&#039;&#039; generally means &#039;&#039;&#039;full-range, flat-response&#039;&#039;&#039;.<br /> <br /> The goal of an FRFR playback system is to reproduce the complete modeled signal, including the cabinet and microphone response, without deliberately adding the strong coloration associated with a traditional guitar cabinet.<br /> <br /> In practice, no loudspeaker system is perfectly flat, and room acoustics also affect the result.<br /> <br /> Nevertheless, the distinction is useful:<br /> <br /> &#039;&#039;&#039;Traditional guitar cabinet:&#039;&#039;&#039; provides its own strong cabinet/speaker character.<br /> <br /> &#039;&#039;&#039;FRFR/full-range system:&#039;&#039;&#039; intended to reproduce the cabinet/microphone character already contained in the model or IR.<br /> <br /> == Cabinet modeling ==<br /> <br /> Not every digital cabinet system uses static IR files in the same way.<br /> <br /> Some hardware and software provides &#039;&#039;&#039;cabinet modeling&#039;&#039;&#039; with controls for variables such as:<br /> <br /> * Cabinet<br /> * Speaker<br /> * Microphone<br /> * Microphone position<br /> * Distance<br /> * Angle<br /> * Resonance<br /> <br /> The implementation may involve IRs, interpolation between measurements, algorithmic processing, more advanced modeling, or combinations of techniques.<br /> <br /> Therefore, the presence of adjustable cabinet controls does not necessarily mean that a single conventional IR is simply being modified.<br /> <br /> == Capturing an amplifier with an IR in the chain ==<br /> <br /> It is technically possible for an IR or cabinet simulation to be active while creating a capture.<br /> <br /> If that processing is present in the measured return path, its effect may become part of the resulting model.<br /> <br /> This can be intentional.<br /> <br /> It can also happen accidentally.<br /> <br /> Before capturing, verify whether any of the following are active:<br /> <br /> * Hardware cabinet simulation<br /> * Load-box cabinet processing<br /> * DAW cabinet plugin<br /> * Interface DSP<br /> * IR loader<br /> * Amplifier direct-output speaker simulation<br /> <br /> If the objective is a direct amplifier capture, these should normally be disabled or bypassed.<br /> <br /> == Identifying what a capture contains ==<br /> <br /> Before adding an IR to a downloaded capture, determine whether cabinet processing is already included.<br /> <br /> Useful clues may come from:<br /> <br /> * Capture metadata<br /> * Filename<br /> * Creator notes<br /> * Platform tags<br /> * Documentation<br /> * Listening tests<br /> <br /> However, filenames alone should not be considered definitive.<br /> <br /> A model named after an amplifier may be an amplifier-only capture or a complete amplifier/cabinet/microphone capture.<br /> <br /> Good capture documentation should state this explicitly.<br /> <br /> == Choosing an IR ==<br /> <br /> There is no universally correct cabinet IR for an amplifier capture.<br /> <br /> Choice depends on the desired result.<br /> <br /> Useful considerations include:<br /> <br /> * Cabinet type<br /> * Speaker<br /> * Microphone<br /> * Microphone position<br /> * Musical style<br /> * Instrument<br /> * Pickup type<br /> * Mix context<br /> * Playback system<br /> <br /> An IR that sounds impressive in isolation may not be the best choice in a mix.<br /> <br /> Likewise, a relatively bright or mid-focused IR that seems less impressive alone may work extremely well with other instruments.<br /> <br /> Evaluate IRs in the context in which they will actually be used.<br /> <br /> == Organizing IR libraries ==<br /> <br /> IR collections can become very large.<br /> <br /> Useful organization may include:<br /> <br /> * Manufacturer<br /> * Cabinet<br /> * Speaker<br /> * Microphone<br /> * Position<br /> * Mix<br /> * Sample rate<br /> * Personal favorites<br /> <br /> Do not assume that owning thousands of IRs necessarily improves results.<br /> <br /> A smaller, well-understood collection can be considerably easier to use than a huge library of poorly documented files.<br /> <br /> == Documenting captures and IRs ==<br /> <br /> When sharing a capture that includes a physical cabinet, useful information includes:<br /> <br /> * Cabinet manufacturer and model<br /> * Speaker manufacturer and model<br /> * Cabinet configuration<br /> * Microphone<br /> * Microphone position<br /> * Distance<br /> * Microphone preamp<br /> * Whether multiple microphones were used<br /> <br /> When sharing a direct capture, state clearly that the cabinet is &#039;&#039;&#039;not&#039;&#039;&#039; included.<br /> <br /> This single piece of information can prevent a great deal of confusion.<br /> <br /> == Practical rule ==<br /> <br /> Before adding cabinet processing, ask:<br /> <br /> &#039;&#039;&#039;Does this capture already contain a cabinet and microphone response?&#039;&#039;&#039;<br /> <br /> If the answer is yes, additional cabinet processing is normally unnecessary.<br /> <br /> If the answer is no, a cabinet IR, cabinet model, or physical cabinet will generally be required for conventional guitar-amplifier playback.<br /> <br /> Knowing what is already present in the model is more important than following a particular preset or signal-chain convention.<br /> <br /> == See also ==<br /> <br /> * [[Capture Types]]<br /> * [[Capture Signal Chain]]<br /> * [[Creating a Capture]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Troubleshooting Captures]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss cabinets, speakers, microphones, and IRs on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Cabinets and IRs]]<br /> [[Category:Capture techniques]]</div> Tue, 01 Sep 2026 05:54:25 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Cabinets_and_IRs https://namforum.com/wiki/index.php?title=Capture_Types&diff=8&oldid=0 https://namforum.com/wiki/index.php?title=Capture_Types&diff=8&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Capture Types =<br /> <br /> Not all captures represent the same portion of an audio signal chain.<br /> <br /> A model may represent a single pedal, an amplifier without a speaker cabinet, an amplifier together with a cabinet and microphone, or an entire chain of equipment.<br /> <br /> Understanding &#039;&#039;&#039;what is contained in a capture&#039;&#039;&#039; is essential. It determines how the model should be used, what additional processing may be required, and whether two models can be meaningfully compared.<br /> <br /> Terminology varies between capture platforms and communities, so the most useful description of a capture is often a clear statement of what equipment was included in the capture signal path.<br /> <br /> == Direct amplifier captures ==<br /> <br /> A &#039;&#039;&#039;direct amplifier capture&#039;&#039;&#039; represents an amplifier without the acoustic contribution of a physical loudspeaker cabinet and microphone.<br /> <br /> A typical signal path is:<br /> <br /> &#039;&#039;&#039;Capture signal → amplifier → suitable load/direct capture device → capture system&#039;&#039;&#039;<br /> <br /> The resulting model can then be used with:<br /> <br /> * A cabinet impulse response (IR)<br /> * A modeled cabinet<br /> * A physical guitar cabinet with an appropriate power-amplification system<br /> * Other compatible speaker/cabinet processing<br /> <br /> Direct captures are useful because the amplifier and cabinet portions of the signal chain remain separate.<br /> <br /> This allows the same amplifier capture to be paired with different cabinets, speakers, microphones, or IRs.<br /> <br /> &#039;&#039;&#039;A speaker-level amplifier output must not be connected directly to a normal audio-interface input unless the equipment is specifically designed to accept that signal.&#039;&#039;&#039;<br /> <br /> See [[Cabinets and IRs]].<br /> <br /> == Amplifier and cabinet captures ==<br /> <br /> A capture can include the amplifier and the response of a loudspeaker cabinet.<br /> <br /> A common microphone-based signal path is:<br /> <br /> &#039;&#039;&#039;Capture signal → amplifier → loudspeaker cabinet → microphone → microphone preamp/interface → capture system&#039;&#039;&#039;<br /> <br /> In this case, the resulting model may represent the combined behavior of:<br /> <br /> * Amplifier<br /> * Loudspeaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Microphone preamp<br /> * Other equipment in the return path<br /> <br /> Such a capture normally does not require an additional cabinet IR during playback unless another cabinet response is intentionally being added.<br /> <br /> Adding a conventional cabinet simulation to a capture that already contains a cabinet and microphone response can produce an unintended &quot;double cabinet&quot; effect.<br /> <br /> == Pedal captures ==<br /> <br /> Many nonlinear pedals can be captured.<br /> <br /> Examples include:<br /> <br /> * Overdrive<br /> * Distortion<br /> * Fuzz<br /> * Boost<br /> * Preamp pedals<br /> * Some analog processors<br /> <br /> A typical chain is:<br /> <br /> &#039;&#039;&#039;Capture signal → pedal → capture system&#039;&#039;&#039;<br /> <br /> or:<br /> <br /> &#039;&#039;&#039;Capture signal → reamp/capture device → pedal → capture system&#039;&#039;&#039;<br /> <br /> The resulting model represents the pedal at the settings used during capture.<br /> <br /> A pedal capture can then be placed before an amplifier model or another compatible part of a virtual signal chain.<br /> <br /> == Pedal and amplifier captures ==<br /> <br /> A pedal and amplifier can be captured together:<br /> <br /> &#039;&#039;&#039;Capture signal → pedal → amplifier → direct return&#039;&#039;&#039;<br /> <br /> or, when a physical cabinet and microphone are included:<br /> <br /> &#039;&#039;&#039;Capture signal → pedal → amplifier → cabinet → microphone → capture system&#039;&#039;&#039;<br /> <br /> This can reproduce a particular combination very effectively because the interaction between the devices is included in the measured system.<br /> <br /> The tradeoff is flexibility.<br /> <br /> If the pedal and amplifier are captured together, they cannot normally be separated afterward. Changing the virtual amplifier while retaining only the captured pedal is not equivalent to having captured the pedal separately.<br /> <br /> == Preamp captures ==<br /> <br /> Preamps and preamp stages can also be capture targets.<br /> <br /> Examples may include:<br /> <br /> * Guitar amplifier preamps<br /> * Rack preamps<br /> * Pedal-format preamps<br /> * Studio preamplifiers<br /> * Other nonlinear line-level equipment<br /> <br /> Whether a particular device can be modeled accurately depends on the capabilities and limitations of the capture technology.<br /> <br /> A preamp capture may require additional processing or amplification during playback depending on its intended use.<br /> <br /> == Complete signal-chain captures ==<br /> <br /> Multiple devices can be captured as one system.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Pedal → preamp → amplifier → cabinet → microphone → microphone preamp&#039;&#039;&#039;<br /> <br /> The resulting model represents the combined behavior of everything included in the measured path.<br /> <br /> This approach can be useful when the objective is to reproduce a specific finished sound rather than provide maximum flexibility.<br /> <br /> It can also preserve interactions between devices that might not be reproduced identically by combining independently captured models.<br /> <br /> The disadvantage is that the individual components cannot normally be adjusted, removed, or rearranged after capture.<br /> <br /> == Captures with microphones ==<br /> <br /> When a microphone is part of the return path, its characteristics become part of the resulting measurement.<br /> <br /> This includes:<br /> <br /> * Microphone model<br /> * Polar pattern<br /> * Position<br /> * Distance<br /> * Angle<br /> * Proximity effect<br /> * Microphone preamp<br /> * Room contribution where significant<br /> <br /> A small change in microphone position can produce a substantial change in the resulting sound.<br /> <br /> Therefore, microphone information is valuable metadata for any capture containing a physical cabinet.<br /> <br /> == Captures without microphones ==<br /> <br /> Direct captures avoid the microphone and acoustic cabinet path.<br /> <br /> They can provide greater flexibility because cabinet and microphone characteristics can be selected later using an IR or cabinet-modeling system.<br /> <br /> This can also make direct captures particularly useful for comparison and experimentation.<br /> <br /> However, the playback chain must then supply whatever cabinet/speaker behavior is required for the intended sound.<br /> <br /> == Full-rig captures ==<br /> <br /> The term &#039;&#039;&#039;full rig&#039;&#039;&#039; is sometimes used for a capture containing most or all of a finished guitar signal chain.<br /> <br /> Depending on the platform and user, this might include:<br /> <br /> * Pedals<br /> * Amplifier<br /> * Cabinet<br /> * Microphone<br /> * Microphone preamp<br /> <br /> Because terminology varies, a label such as &quot;full rig&quot; should not replace actual documentation of what the capture contains.<br /> <br /> When sharing a capture, list the components explicitly whenever possible.<br /> <br /> == Static settings and multiple captures ==<br /> <br /> Most capture technologies reproduce the behavior of equipment around the configuration used during capture.<br /> <br /> For example, an amplifier captured with:<br /> <br /> * Gain at 4<br /> * Bass at 5<br /> * Middle at 6<br /> * Treble at 5<br /> * Presence at 4<br /> <br /> does not automatically become a complete digital implementation of that amplifier&#039;s control circuitry.<br /> <br /> Playback software may provide controls that modify the resulting model, but those controls are not necessarily equivalent to turning the corresponding controls on the original device.<br /> <br /> For substantially different settings, multiple captures may be appropriate.<br /> <br /> A collection might therefore include:<br /> <br /> * Clean<br /> * Edge of breakup<br /> * Crunch<br /> * High gain<br /> * Lead<br /> <br /> or several captures made at documented control settings.<br /> <br /> == Linear and nonlinear elements ==<br /> <br /> Capture systems are particularly useful for modeling &#039;&#039;&#039;nonlinear&#039;&#039;&#039; behavior such as distortion, saturation, compression, and level-dependent response.<br /> <br /> Some portions of an audio chain, particularly loudspeaker cabinet and microphone responses, can also be represented using linear techniques such as [[Cabinets and IRs|impulse responses]].<br /> <br /> This creates several possible workflows.<br /> <br /> For example:<br /> <br /> &#039;&#039;&#039;Captured amplifier + separate cabinet IR&#039;&#039;&#039;<br /> <br /> provides more flexibility than:<br /> <br /> &#039;&#039;&#039;Captured amplifier + cabinet + microphone&#039;&#039;&#039;<br /> <br /> while the latter may conveniently reproduce a specific complete sound.<br /> <br /> Neither approach is inherently superior. They serve different purposes.<br /> <br /> == What a capture does not necessarily contain ==<br /> <br /> Do not infer the contents of a capture solely from its filename.<br /> <br /> For example, a model named after an amplifier might contain:<br /> <br /> * Amplifier only<br /> * Amplifier and cabinet<br /> * Amplifier, cabinet, and microphone<br /> * Pedal and amplifier<br /> * An entire signal chain<br /> <br /> Likewise, the word &quot;direct&quot; may be used differently by different creators and platforms.<br /> <br /> When the information is available, consult the capture&#039;s metadata or creator documentation.<br /> <br /> == Choosing a capture type ==<br /> <br /> The appropriate capture type depends on the intended use.<br /> <br /> Choose a &#039;&#039;&#039;direct amplifier capture&#039;&#039;&#039; when you want to experiment with different cabinets, IRs, or playback systems.<br /> <br /> Choose an &#039;&#039;&#039;amplifier/cabinet capture&#039;&#039;&#039; when you want a particular miked cabinet sound already included.<br /> <br /> Choose a &#039;&#039;&#039;pedal capture&#039;&#039;&#039; when you want to combine that pedal behavior with different amplifier models.<br /> <br /> Choose a &#039;&#039;&#039;combined or full-chain capture&#039;&#039;&#039; when reproducing a particular complete signal chain is more important than independently changing its components.<br /> <br /> There is no universally best capture type.<br /> <br /> The important question is:<br /> <br /> &#039;&#039;&#039;What part of the original signal chain do you want the model to reproduce?&#039;&#039;&#039;<br /> <br /> == Documenting capture type ==<br /> <br /> When sharing captures, state clearly what they contain.<br /> <br /> Useful information includes:<br /> <br /> * Capture type<br /> * Devices included<br /> * Devices deliberately excluded<br /> * Whether a cabinet is included<br /> * Whether a microphone is included<br /> * Cabinet and speaker<br /> * Microphone and position<br /> * Pedal settings<br /> * Amplifier settings<br /> * Capture method<br /> * Calibration information<br /> <br /> Clear documentation prevents incorrect playback configurations and makes captures more useful to other people.<br /> <br /> == See also ==<br /> <br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Creating a Capture]]<br /> * [[Cabinets and IRs]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Troubleshooting Captures]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss capture types and signal chains on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Capture techniques]]</div> Tue, 01 Sep 2026 05:53:17 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Capture_Types https://namforum.com/wiki/index.php?title=Creating_a_Capture&diff=7&oldid=0 https://namforum.com/wiki/index.php?title=Creating_a_Capture&diff=7&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Creating a Capture =<br /> <br /> Creating a reliable capture is a controlled measurement process.<br /> <br /> The exact procedure differs between [[Neural Amp Modeler]], [[TONEX]], [[Kemper Profiling]], [[Neural Capture]], [[Line 6 Proxy]], and other systems, but the underlying principles are similar: establish a known signal path, configure the equipment being captured, send the required test or training signal through it, record or measure the response, and evaluate the resulting model.<br /> <br /> This article describes the general workflow. Always follow the platform-specific instructions for the capture system being used.<br /> <br /> == 1. Decide what you are capturing ==<br /> <br /> Before connecting equipment, define exactly what the resulting model is intended to represent.<br /> <br /> Possible capture targets include:<br /> <br /> * A pedal<br /> * A preamp<br /> * An amplifier without a cabinet<br /> * An amplifier and cabinet<br /> * An amplifier, cabinet, and microphone<br /> * Multiple pedals or processors<br /> * A complete signal chain<br /> <br /> This decision determines both the capture signal path and what additional processing may be required during playback.<br /> <br /> For example, a direct amplifier capture may require a cabinet or impulse response during playback, while a capture made through a physical cabinet and microphone may already contain the cabinet and microphone characteristics.<br /> <br /> See [[Capture Types]] and [[Cabinets and IRs]].<br /> <br /> == 2. Configure the device under test ==<br /> <br /> Set the equipment to the configuration you want to capture.<br /> <br /> Document relevant controls and settings before beginning.<br /> <br /> Depending on the device, these may include:<br /> <br /> * Channel<br /> * Gain<br /> * Volume<br /> * Tone controls<br /> * Presence or resonance<br /> * Bright, boost, or voicing switches<br /> * Pedal controls<br /> * Cabinet configuration<br /> * Speaker selection<br /> * Microphone and position<br /> * Power or operating mode<br /> * Other devices in the chain<br /> <br /> A capture normally represents the behavior of the equipment at a particular configuration.<br /> <br /> If several substantially different sounds are required, it may be better to create several captures rather than expect one capture to represent every possible setting.<br /> <br /> == 3. Allow equipment to stabilize ==<br /> <br /> Equipment should be in its normal operating condition before critical captures are made.<br /> <br /> Tube amplifiers in particular should be allowed to warm up and stabilize.<br /> <br /> Also verify that:<br /> <br /> * Batteries or power supplies are reliable<br /> * Pedals are operating normally<br /> * Amplifier operating modes will not change<br /> * Noise gates or automatic processes are understood<br /> * Nothing is likely to switch state during the capture<br /> <br /> A changing device produces changing measurement data.<br /> <br /> == 4. Build the capture signal chain ==<br /> <br /> Establish the complete path before running the capture.<br /> <br /> A common direct chain is:<br /> <br /> &#039;&#039;&#039;Capture/training signal → audio interface output → reamp or capture device → device under test → audio interface input → capture system&#039;&#039;&#039;<br /> <br /> A microphone-based amplifier capture may instead use:<br /> <br /> &#039;&#039;&#039;Capture/training signal → interface → reamp/capture device → amplifier → cabinet → microphone → preamp/interface → capture system&#039;&#039;&#039;<br /> <br /> A direct amplifier capture may require a suitable load or dedicated capture device between the amplifier&#039;s speaker output and the recording interface.<br /> <br /> See [[Capture Signal Chain]].<br /> <br /> &#039;&#039;&#039;Never connect an amplifier speaker output directly to an ordinary audio-interface input unless the equipment is specifically designed to accept that signal.&#039;&#039;&#039;<br /> <br /> == 5. Verify routing ==<br /> <br /> Before running the capture, verify each part of the signal path individually.<br /> <br /> Confirm that:<br /> <br /> # The intended output carries the test signal.<br /> # The signal reaches the device under test.<br /> # The device responds normally.<br /> # The return reaches the intended interface input.<br /> # The correct input is being recorded.<br /> # Monitoring does not create a feedback loop.<br /> # No unintended processing is active.<br /> <br /> Do not rely only on software meters. Where practical, verify that the physical equipment itself is receiving and returning the expected signal.<br /> <br /> == 6. Establish gain staging ==<br /> <br /> Check the level at every important stage of the chain.<br /> <br /> The objective is to avoid unintended clipping while maintaining a useful signal-to-noise ratio.<br /> <br /> Potential overload points include:<br /> <br /> * Interface output<br /> * Reamp or capture device<br /> * Pedal or processor stages<br /> * Amplifier stages<br /> * Load/direct-output device<br /> * Microphone preamp<br /> * Interface input<br /> * Digital recording path<br /> <br /> Remember that a return signal below 0 dBFS does not prove that an earlier analog stage has not clipped.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == 7. Calibrate where required ==<br /> <br /> If the capture platform provides a calibration procedure, perform it before making the capture.<br /> <br /> Calibration may establish the relationship between:<br /> <br /> * Digital test-signal level<br /> * Interface output voltage<br /> * Signal presented to the device under test<br /> * Return level<br /> * Model input or output level<br /> <br /> Do not substitute an arbitrary level simply because it produces a good-looking waveform.<br /> <br /> A capture can sound convincing while still having incorrect level calibration.<br /> <br /> Platform-specific calibration requirements should take precedence over generic advice.<br /> <br /> == 8. Remove unintended processing ==<br /> <br /> Unless specifically required by the capture system, avoid modifying the test or return signal.<br /> <br /> Check for:<br /> <br /> * EQ<br /> * Compression<br /> * Limiting<br /> * Noise reduction<br /> * Normalization<br /> * Cabinet simulation<br /> * Speaker simulation<br /> * DAW plugins<br /> * Interface DSP<br /> * Automatic gain control<br /> * Sample-rate conversion<br /> * Unintended fades<br /> <br /> Any processing included in the measurement path can become part of the behavior the capture system attempts to reproduce.<br /> <br /> If processing is intentionally part of the target signal chain, document it.<br /> <br /> == 9. Run the capture or record the training response ==<br /> <br /> Follow the procedure required by the chosen platform.<br /> <br /> Some systems perform the measurement and model creation as part of an integrated capture process.<br /> <br /> Other systems, including some [[Neural Amp Modeler|NAM]] workflows, may use a training signal that is played through the device and recorded. The resulting recording is then used in a separate training process.<br /> <br /> During the measurement:<br /> <br /> * Do not change controls.<br /> * Do not switch channels.<br /> * Avoid touching cables unnecessarily.<br /> * Avoid introducing unrelated sounds into microphone-based captures.<br /> * Watch for clipping or unexpected level changes.<br /> * Allow the process to complete normally.<br /> <br /> If the procedure fails, identify the cause before repeatedly running additional captures.<br /> <br /> == 10. Preserve the original recording when applicable ==<br /> <br /> When a platform uses a recorded training response, retain an unprocessed copy of the original recording.<br /> <br /> Do not permanently overwrite the only copy with:<br /> <br /> * Normalization<br /> * EQ<br /> * Noise reduction<br /> * Resampling<br /> * Trimming<br /> * Format conversion<br /> <br /> If editing is required by the training procedure, work from a copy.<br /> <br /> The original recording can be valuable for troubleshooting, retraining, comparison, or future training methods.<br /> <br /> == 11. Create or train the model ==<br /> <br /> Once the measurement or recording is complete, the capture system creates or trains the resulting model.<br /> <br /> Depending on the technology, this may occur:<br /> <br /> * Locally on a computer<br /> * Inside dedicated hardware<br /> * Through manufacturer software<br /> * Through an online or cloud service<br /> * Through third-party training software<br /> <br /> Training time can vary substantially according to the platform, model architecture, computer hardware, and selected quality settings.<br /> <br /> A process completing successfully means that a model was created. It does not necessarily mean that the model is accurate.<br /> <br /> == 12. Compare the capture with the original ==<br /> <br /> Evaluation is an essential part of the process.<br /> <br /> Where practical, compare the model and the original device using the same source signal.<br /> <br /> Listen for differences in:<br /> <br /> * Overall gain<br /> * Distortion character<br /> * Dynamic response<br /> * Attack<br /> * Compression<br /> * Low-frequency behavior<br /> * High-frequency behavior<br /> * Noise<br /> * Sustain<br /> * Response to different playing levels<br /> <br /> Level-match the comparison carefully. A louder signal is easily perceived as more impressive even when it is not more accurate.<br /> <br /> For critical evaluation, switching between the original and the model while using the same source material can reveal differences that are difficult to identify from memory.<br /> <br /> == 13. Test more than one source ==<br /> <br /> A model that sounds accurate with the training or comparison material may behave differently with another source.<br /> <br /> Where appropriate, test with:<br /> <br /> * Different guitars<br /> * Different pickup types<br /> * Bass<br /> * Different playing dynamics<br /> * Single notes<br /> * Chords<br /> * Clean playing<br /> * Hard transients<br /> * Sustained notes<br /> <br /> For captures intended for non-guitar sources, test them with representative material from those sources.<br /> <br /> The objective is to determine whether the model reproduces the useful behavior of the original equipment, not merely whether one example sounds similar.<br /> <br /> == 14. Name the capture meaningfully ==<br /> <br /> A large collection quickly becomes difficult to manage if captures have ambiguous names.<br /> <br /> A useful naming convention may identify:<br /> <br /> * Manufacturer<br /> * Device<br /> * Channel or mode<br /> * Gain setting<br /> * Cabinet status<br /> * Microphone where applicable<br /> * Significant pedal or switch settings<br /> <br /> For example, a collection might distinguish clearly between several captures of the same amplifier at different gain or channel settings.<br /> <br /> Avoid names that only make sense to the person who created them.<br /> <br /> == 15. Document the capture ==<br /> <br /> Good metadata increases the value of a capture.<br /> <br /> Useful information can include:<br /> <br /> * Equipment manufacturer and model<br /> * Device settings<br /> * Capture date<br /> * Capture platform and software version<br /> * Audio interface<br /> * Reamp or capture device<br /> * Calibration information<br /> * Cabinet and speaker<br /> * Microphone and position<br /> * Microphone preamp<br /> * Sample rate<br /> * Signal-chain details<br /> * Whether a cabinet is included<br /> * Known limitations or unusual behavior<br /> <br /> If the capture is shared publicly, this information helps other users understand what they are loading.<br /> <br /> == 16. Keep unsuccessful captures ==<br /> <br /> Not every unsuccessful experiment should immediately be deleted.<br /> <br /> When practical, preserve enough information to understand why a capture failed or produced an unexpected result.<br /> <br /> A failed capture can reveal useful information about:<br /> <br /> * Calibration<br /> * Signal routing<br /> * Equipment behavior<br /> * Training settings<br /> * Platform limitations<br /> * Noise<br /> * Clipping<br /> * Model architecture<br /> <br /> Controlled failures can be useful experiments.<br /> <br /> == 17. Share reproducible information ==<br /> <br /> When discussing a capture with other users, provide enough information for the problem or result to be understood.<br /> <br /> Instead of simply reporting:<br /> <br /> &#039;&#039;&quot;This capture sounds wrong.&quot;&#039;&#039;<br /> <br /> provide information such as:<br /> <br /> * What was captured<br /> * Signal chain<br /> * Interface<br /> * Reamp method<br /> * Calibration<br /> * Device settings<br /> * Capture platform and version<br /> * Playback system<br /> * What differs from the original<br /> <br /> This turns an individual experience into information that others can investigate and learn from.<br /> <br /> == A repeatable workflow ==<br /> <br /> A useful general workflow is:<br /> <br /> # Define the capture target.<br /> # Document the equipment settings.<br /> # Build the signal chain.<br /> # Verify routing.<br /> # Establish gain staging.<br /> # Calibrate where required.<br /> # Run the capture.<br /> # Preserve the source measurement where applicable.<br /> # Create or train the model.<br /> # Compare it with the original.<br /> # Test it with varied material.<br /> # Document the result.<br /> # Repeat only the variables you intend to change.<br /> <br /> The objective is not merely to produce a model file.<br /> <br /> The objective is to produce a model whose origin, behavior, and limitations are understood.<br /> <br /> == See also ==<br /> <br /> * [[Getting Started with Capture Technology]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss capture techniques and results on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Capture techniques]]</div> Tue, 01 Sep 2026 05:52:06 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Creating_a_Capture https://namforum.com/wiki/index.php?title=Reamping_for_Capture&diff=6&oldid=0 https://namforum.com/wiki/index.php?title=Reamping_for_Capture&diff=6&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Reamping for Capture =<br /> <br /> &#039;&#039;&#039;Reamping&#039;&#039;&#039; is the process of sending a previously recorded or generated audio signal from a recording system into an amplifier, pedal, preamp, or other piece of audio equipment.<br /> <br /> The technique originated as a studio-production method, allowing a performance recorded earlier to be played through different equipment later. Capture technology uses a closely related process: a known test or training signal is sent from the capture system through the equipment being modeled and the result is recorded.<br /> <br /> Reamping is therefore an important part of many capture workflows.<br /> <br /> == Basic reamping path ==<br /> <br /> A typical capture path using a reamp device is:<br /> <br /> &#039;&#039;&#039;Capture or training signal → audio interface line output → reamp device → device under test → return path → audio interface input&#039;&#039;&#039;<br /> <br /> The reamp device sits between the audio interface output and equipment that expects an instrument-type input.<br /> <br /> See [[Capture Signal Chain]] for the complete signal path.<br /> <br /> == Why use a reamp device? ==<br /> <br /> An audio interface normally provides a line-level output, often balanced and relatively low impedance.<br /> <br /> A guitar amplifier or pedal is commonly designed to receive an unbalanced instrument signal.<br /> <br /> A reamp device can provide a convenient interface between these environments.<br /> <br /> Depending on its design, a reamp device may provide:<br /> <br /> * Balanced-to-unbalanced conversion<br /> * Signal attenuation or level adjustment<br /> * Transformer isolation<br /> * Ground-lift capability<br /> * Appropriate connectors<br /> * Convenient routing to instrument equipment<br /> <br /> Not every reamp device provides all of these functions.<br /> <br /> == Is impedance matching required? ==<br /> <br /> Reamping is often described simply as &quot;impedance matching,&quot; but that description can be misleading.<br /> <br /> In many practical systems, the important functions of a reamp device are &#039;&#039;&#039;level conversion, balanced-to-unbalanced interfacing, isolation, grounding, and predictable connection&#039;&#039;&#039; rather than literal impedance matching in the traditional maximum-power-transfer sense.<br /> <br /> Modern audio equipment is generally designed using voltage bridging: a relatively low source impedance drives a substantially higher input impedance.<br /> <br /> The relevant question is therefore not merely whether two impedance numbers &quot;match,&quot; but whether the source and destination behave appropriately when connected.<br /> <br /> Some amplifiers, pedals, and especially certain vintage or unusual circuits may respond differently to source impedance, so the requirements of the actual equipment should still be considered.<br /> <br /> == Passive and active reamp devices ==<br /> <br /> Reamp devices can be passive or active.<br /> <br /> === Passive reamp devices ===<br /> <br /> Passive designs commonly use a transformer and attenuation network.<br /> <br /> Potential advantages include:<br /> <br /> * Electrical isolation<br /> * Simple operation<br /> * No external power requirement<br /> * Ground-loop reduction<br /> <br /> Their available output depends on the level supplied by the interface and the design of the device.<br /> <br /> === Active reamp devices ===<br /> <br /> Active designs use powered electronics and may provide:<br /> <br /> * Additional level control<br /> * Buffering<br /> * Greater available output<br /> * Different impedance characteristics<br /> * Additional routing features<br /> <br /> Neither type is automatically superior for capture work. The appropriate choice depends on the interface, equipment being captured, calibration requirements, and desired workflow.<br /> <br /> == Level is critical ==<br /> <br /> A reamp device can change the level reaching the device under test.<br /> <br /> This is especially important when capturing nonlinear equipment.<br /> <br /> For example, increasing the signal sent into an overdrive pedal or amplifier may produce more saturation even though none of the controls on the device itself have changed.<br /> <br /> Therefore, the reamp output setting is part of the capture configuration.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Reamp level controls ==<br /> <br /> If the reamp device has an adjustable output control, document its setting.<br /> <br /> However, markings such as:<br /> <br /> * 12 o&#039;clock<br /> * 50%<br /> * -10 dB<br /> * Unity<br /> <br /> should not automatically be assumed to represent a standardized instrument level.<br /> <br /> If accurate calibration is required, measure or otherwise establish the actual relationship between the digital test signal and the electrical output.<br /> <br /> == Ground loops ==<br /> <br /> Capture setups can be particularly susceptible to ground loops because they often connect several mains-powered devices simultaneously.<br /> <br /> A typical setup may include:<br /> <br /> * Computer<br /> * Audio interface<br /> * Amplifier<br /> * Powered capture equipment<br /> * Monitor system<br /> <br /> A transformer-isolated reamp output can break an unwanted audio-frequency ground path while still allowing the signal to pass.<br /> <br /> Some reamp devices also provide a ground-lift switch.<br /> <br /> &#039;&#039;&#039;Never defeat the protective mains safety earth of an amplifier or other equipment in an attempt to eliminate hum.&#039;&#039;&#039;<br /> <br /> Solve grounding problems in the signal path using equipment and methods intended for that purpose.<br /> <br /> == Balanced and unbalanced connections ==<br /> <br /> Professional audio-interface outputs are frequently balanced, while guitar amplifiers and pedals normally use unbalanced TS connections.<br /> <br /> A reamp device commonly accepts the balanced signal from the interface and provides an unbalanced instrument connection at its output.<br /> <br /> This can also make longer cable runs between the interface and reamp device less susceptible to interference, since the balanced portion of the connection can extend closer to the amplifier or pedal.<br /> <br /> == Direct connection without a reamp device ==<br /> <br /> Some interfaces and capture devices provide outputs specifically designed for reamping or instrument-level operation.<br /> <br /> In other situations, an ordinary interface line output may be usable directly with the equipment being captured.<br /> <br /> Therefore:<br /> <br /> &#039;&#039;&#039;A reamp box is not automatically mandatory simply because a capture is being made.&#039;&#039;&#039;<br /> <br /> The decision depends on:<br /> <br /> * Available interface outputs<br /> * Required signal level<br /> * Source and input impedance<br /> * Balanced/unbalanced connection requirements<br /> * Grounding and isolation<br /> * Platform calibration requirements<br /> * Equipment being captured<br /> <br /> If a direct connection provides the required electrical conditions and calibration, adding another device solely because it is called a reamp box may provide no benefit.<br /> <br /> == Reamping pedals ==<br /> <br /> A common pedal capture chain is:<br /> <br /> &#039;&#039;&#039;Interface output → reamp device → pedal → interface input&#039;&#039;&#039;<br /> <br /> For multiple pedals:<br /> <br /> &#039;&#039;&#039;Interface output → reamp device → pedal 1 → pedal 2 → interface input&#039;&#039;&#039;<br /> <br /> The resulting capture represents the combined behavior of all devices included in the chain.<br /> <br /> Pay attention to bypass state, switching, power supply, input level, output level, and any time-varying behavior.<br /> <br /> == Reamping amplifiers ==<br /> <br /> For an amplifier capture, the reamp output normally feeds the amplifier&#039;s instrument input.<br /> <br /> What happens after the amplifier depends on the type of capture.<br /> <br /> For a microphone-based capture:<br /> <br /> &#039;&#039;&#039;Interface → reamp → amplifier → cabinet → microphone → preamp/interface&#039;&#039;&#039;<br /> <br /> For a direct amplifier capture:<br /> <br /> &#039;&#039;&#039;Interface → reamp → amplifier → suitable load/direct capture device → interface&#039;&#039;&#039;<br /> <br /> &#039;&#039;&#039;Never connect a speaker-level amplifier output directly to a normal audio-interface input unless the equipment is specifically designed to permit that connection.&#039;&#039;&#039;<br /> <br /> See [[Capture Types]] and [[Cabinets and IRs]].<br /> <br /> == Using a DAW ==<br /> <br /> When a DAW is used to play a capture training file and record the result, verify the routing carefully.<br /> <br /> A simple configuration may use:<br /> <br /> * One track containing the training signal<br /> * One dedicated interface output<br /> * One return input<br /> * One recording track<br /> <br /> Avoid processing the training or return tracks unless the capture procedure specifically requires it.<br /> <br /> Check for:<br /> <br /> * Plugins<br /> * Fader offsets<br /> * Pan laws<br /> * Sends<br /> * Bus processing<br /> * Automatic normalization<br /> * Sample-rate conversion<br /> * Monitoring loops<br /> <br /> The objective is usually to preserve the test and return signals as predictably as possible.<br /> <br /> == Record the configuration ==<br /> <br /> For repeatable captures, document:<br /> <br /> * Interface<br /> * Interface output<br /> * Output operating level<br /> * Reamp device<br /> * Reamp output setting<br /> * Device under test<br /> * Device settings<br /> * Return connection<br /> * Interface input<br /> * Input gain<br /> * Calibration procedure<br /> * Capture platform and version<br /> <br /> If a later capture sounds different, these records can help determine what changed.<br /> <br /> == Reamping as a diagnostic tool ==<br /> <br /> A useful advantage of a reamping-based capture workflow is repeatability.<br /> <br /> The same test signal can be sent through:<br /> <br /> * Different amplifiers<br /> * Different pedals<br /> * Different settings<br /> * Different reamp devices<br /> * Different calibration levels<br /> <br /> This makes controlled comparisons possible.<br /> <br /> When investigating a capture problem, change one variable at a time and compare the results.<br /> <br /> == See also ==<br /> <br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Troubleshooting Captures]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss reamping and capture setups on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Capture techniques]]<br /> [[Category:Reamping]]</div> Tue, 01 Sep 2026 05:50:11 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Reamping_for_Capture https://namforum.com/wiki/index.php?title=Gain_Staging_and_Calibration&diff=5&oldid=0 https://namforum.com/wiki/index.php?title=Gain_Staging_and_Calibration&diff=5&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Gain Staging and Calibration =<br /> <br /> &#039;&#039;&#039;Gain staging&#039;&#039;&#039; and &#039;&#039;&#039;calibration&#039;&#039;&#039; are fundamental to creating accurate and portable captures.<br /> <br /> A capture system does not work only with the sound of a device. It also measures how that device responds to a particular input level. If the level presented to the device during capture is different from the level assumed during playback, the resulting model may respond differently from the original equipment.<br /> <br /> This is especially important with nonlinear equipment such as amplifiers, overdrive pedals, distortion devices, compressors, preamps, and other equipment whose behavior changes with signal level.<br /> <br /> == Gain staging and calibration are not the same thing ==<br /> <br /> The terms are related but describe different concerns.<br /> <br /> &#039;&#039;&#039;Gain staging&#039;&#039;&#039; is the management of signal levels throughout the capture chain so that each stage operates in an appropriate range without unwanted clipping, excessive noise, or unintended attenuation.<br /> <br /> &#039;&#039;&#039;Calibration&#039;&#039;&#039; establishes a meaningful relationship between the digital signal level inside the capture system and the electrical signal level presented to or received from the device under test.<br /> <br /> A capture can have excellent gain staging and still be incorrectly calibrated.<br /> <br /> Likewise, a nominally calibrated system can produce a poor capture if a stage in the signal chain clips or has excessive noise.<br /> <br /> == Why input level matters ==<br /> <br /> Many devices used in music production are nonlinear.<br /> <br /> Consider an amplifier. A small input signal may produce a relatively clean output. Increasing the input level may cause progressively more compression, harmonic distortion, saturation, or clipping.<br /> <br /> A capture made with the wrong input level therefore may not merely be louder or quieter. It may represent different behavior.<br /> <br /> The same principle applies to many pedals, preamps, compressors, and analog processors.<br /> <br /> == Digital level and analog level ==<br /> <br /> Digital audio level is commonly expressed in &#039;&#039;&#039;dBFS&#039;&#039;&#039; (decibels relative to full scale).<br /> <br /> Analog equipment may instead be described using units such as:<br /> <br /> * dBu<br /> * dBV<br /> * volts RMS<br /> * volts peak<br /> * instrument level<br /> * line level<br /> <br /> These are not interchangeable.<br /> <br /> For example, a test signal at -12 dBFS does not by itself tell us the voltage appearing at an audio interface output.<br /> <br /> That voltage depends on the interface, output configuration, reference level, gain controls, and possibly software settings.<br /> <br /> Consequently:<br /> <br /> &#039;&#039;&#039;dBFS alone does not define the electrical level presented to the device under test.&#039;&#039;&#039;<br /> <br /> == Interface output level ==<br /> <br /> Different audio interfaces can produce different analog voltages for the same digital signal.<br /> <br /> Some interfaces also provide selectable operating levels, variable output controls, or different specifications for balanced and unbalanced operation.<br /> <br /> Therefore, two users can play exactly the same digital training file at exactly the same DAW level and still present different electrical levels to identical amplifiers.<br /> <br /> This is one reason calibration becomes important when captures are exchanged between systems.<br /> <br /> == Reamp devices and level ==<br /> <br /> A [[Reamping for Capture|reamp device]] can alter the level between the interface and the device under test.<br /> <br /> Passive and active reamp devices may have different characteristics, and some provide adjustable output controls.<br /> <br /> Changing the reamp level changes the signal reaching the equipment and therefore may change the behavior being captured.<br /> <br /> For repeatable work, document the reamp configuration and any variable controls.<br /> <br /> A reamp device should not be treated as a universal calibration device unless it is specifically designed and used for that purpose.<br /> <br /> == Input calibration ==<br /> <br /> Input calibration concerns the relationship between the digital test signal and the analog level presented to the device being captured.<br /> <br /> The exact procedure depends on the capture platform.<br /> <br /> Where a platform provides an official calibration procedure or reference level, follow that procedure rather than substituting a generic value.<br /> <br /> Useful measurements may involve:<br /> <br /> * Known digital test tones<br /> * Interface output specifications<br /> * AC voltage measurements<br /> * Known reference levels<br /> * Capture-system calibration values<br /> <br /> The objective is not simply to make the signal &quot;loud enough.&quot; The objective is to know what electrical level corresponds to the digital level used by the capture system.<br /> <br /> == Output and return calibration ==<br /> <br /> The return side also matters.<br /> <br /> After passing through the device under test, the signal enters an audio interface or capture device.<br /> <br /> The interface input gain determines the relationship between the analog output of the device and the recorded digital level.<br /> <br /> Incorrect return gain may cause:<br /> <br /> * Clipping<br /> * Excessive noise<br /> * Incorrect model output level<br /> * Poor signal-to-noise ratio<br /> * Difficulty comparing the capture with the original equipment<br /> <br /> Some capture systems normalize or otherwise compensate for portions of the return level. Others may retain meaningful level relationships.<br /> <br /> Consult the documentation for the specific platform.<br /> <br /> == Avoid clipping ==<br /> <br /> Digital clipping occurs when the signal exceeds the available digital range.<br /> <br /> Analog clipping can occur earlier in the chain even when the recorded waveform never reaches 0 dBFS.<br /> <br /> Potential clipping points include:<br /> <br /> * Interface output stage<br /> * Reamp or capture device<br /> * Device under test<br /> * Load or direct-output device<br /> * Microphone preamp<br /> * Interface input stage<br /> * Digital recording path<br /> <br /> A meter showing no digital clipping proves only that the measured digital stage did not exceed full scale. It does not prove that every analog stage was clean.<br /> <br /> Of course, intentional distortion produced by the device being captured is different from unintended overload elsewhere in the signal chain.<br /> <br /> == Signal-to-noise ratio ==<br /> <br /> Setting levels unnecessarily low can create another problem.<br /> <br /> If the return signal is very quiet and later requires substantial amplification, noise from the equipment and recording chain is amplified along with it.<br /> <br /> The objective is therefore not simply &quot;keep everything low.&quot;<br /> <br /> Good gain staging means maintaining appropriate levels throughout the chain while preserving adequate headroom.<br /> <br /> == Controls that can invalidate calibration ==<br /> <br /> Be aware of any variable control between the digital test signal and the device under test.<br /> <br /> Examples include:<br /> <br /> * DAW faders<br /> * Interface mixer levels<br /> * Hardware output knobs<br /> * Reamp level controls<br /> * Digital mixer routing<br /> * Monitor-controller levels<br /> * Software mixer attenuation<br /> <br /> Likewise, variable controls in the return path can change the recorded level.<br /> <br /> Once a system has been calibrated, avoid changing these controls unless the calibration is repeated or the effect of the change is known.<br /> <br /> == Document calibration ==<br /> <br /> Captures intended for distribution should include meaningful calibration information whenever the platform supports it.<br /> <br /> Useful information can include:<br /> <br /> * Audio interface model<br /> * Interface output used<br /> * Output reference level or setting<br /> * Reamp/capture device<br /> * Reamp control position<br /> * Capture-system input calibration value<br /> * Interface input used<br /> * Input gain<br /> * Relevant measurement procedure<br /> * Capture software and version<br /> <br /> This information makes it easier to reproduce, diagnose, and compare captures.<br /> <br /> == Platform-specific requirements ==<br /> <br /> There is no single calibration procedure that should automatically be applied to every capture platform.<br /> <br /> [[Neural Amp Modeler]], [[TONEX]], [[Kemper Profiling]], [[Neural Capture]], [[Line 6 Proxy]], and other systems may use different capture signals, workflows, assumptions, and compensation methods.<br /> <br /> Follow the official requirements of the system being used.<br /> <br /> Platform-specific calibration information belongs in the corresponding platform articles.<br /> <br /> == Practical principle ==<br /> <br /> When troubleshooting a capture, separate two questions:<br /> <br /> # &#039;&#039;&#039;Is the signal clean and appropriately staged?&#039;&#039;&#039;<br /> # &#039;&#039;&#039;Is the signal calibrated to the level expected by the capture system?&#039;&#039;&#039;<br /> <br /> They are not the same question.<br /> <br /> A clean waveform is not proof of correct calibration, and correct calibration is not proof that nothing in the chain is clipping.<br /> <br /> == See also ==<br /> <br /> * [[Capture Signal Chain]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Troubleshooting Captures]]<br /> * [[Capture Types]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss gain staging and calibration on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Capture techniques]]</div> Tue, 01 Sep 2026 05:49:42 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Gain_Staging_and_Calibration https://namforum.com/wiki/index.php?title=Capture_Signal_Chain&diff=4&oldid=0 https://namforum.com/wiki/index.php?title=Capture_Signal_Chain&diff=4&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Capture Signal Chain =<br /> <br /> A &#039;&#039;&#039;capture signal chain&#039;&#039;&#039; is the complete path a test or training signal follows from the capture system, through the equipment being modeled, and back into the recording or capture system.<br /> <br /> Understanding this path is fundamental to creating reliable captures. Many apparent problems with a capture or model are actually problems somewhere in the signal chain.<br /> <br /> == Basic signal flow ==<br /> <br /> A typical direct capture chain is:<br /> <br /> &#039;&#039;&#039;Capture or training signal → audio interface output → reamp/capture device → device under test → audio interface input → capture or recording software&#039;&#039;&#039;<br /> <br /> The exact chain varies according to what is being captured and which capture technology is being used.<br /> <br /> The &#039;&#039;&#039;device under test&#039;&#039;&#039; (DUT) may be an amplifier, pedal, preamp, processor, or a combination of devices.<br /> <br /> == The outgoing signal ==<br /> <br /> Most capture systems begin with a known test or training signal.<br /> <br /> Depending on the platform, this may consist of sweeps, noise, impulses, specially generated audio, or a prerecorded training file.<br /> <br /> The purpose is to present the device under test with a known input so that the system can compare the original signal with the returned signal and determine how the equipment changed it.<br /> <br /> The test signal should normally be used exactly as specified by the capture platform. Changing its level, duration, sample rate, timing, or content without understanding the consequences may invalidate the capture.<br /> <br /> == Audio interface output ==<br /> <br /> The test signal normally leaves the computer through an audio interface output.<br /> <br /> This creates an important electrical issue: an interface line output is not necessarily equivalent to the signal produced by a guitar or other instrument.<br /> <br /> Differences may include:<br /> <br /> * Signal level<br /> * Source impedance<br /> * Balanced versus unbalanced operation<br /> * Grounding<br /> * Connector type<br /> <br /> Simply connecting an interface output directly to an amplifier or pedal may work in some circumstances, but physical compatibility does not necessarily mean electrical equivalence.<br /> <br /> See [[Gain Staging and Calibration]] and [[Reamping for Capture]].<br /> <br /> == Reamp and capture devices ==<br /> <br /> A reamp or capture device may be placed between the audio interface and the device under test.<br /> <br /> Depending on its design, it may provide one or more of the following:<br /> <br /> * Level adjustment<br /> * Balanced-to-unbalanced conversion<br /> * Galvanic isolation<br /> * Ground-lift capability<br /> * Appropriate connection to instrument-level equipment<br /> * Convenient routing<br /> <br /> A reamp box should not automatically be assumed to be necessary for every capture setup. Likewise, the presence of a reamp box does not by itself guarantee correct calibration.<br /> <br /> The requirements depend on the equipment, interface, capture platform, and desired accuracy.<br /> <br /> == Device under test ==<br /> <br /> The device being captured should be placed in a stable, repeatable state before beginning the capture.<br /> <br /> Document settings that may affect the result, including:<br /> <br /> * Gain<br /> * Volume<br /> * Tone controls<br /> * Channel selection<br /> * Switch positions<br /> * Pedal settings<br /> * Cabinet configuration<br /> * Speaker or cabinet simulation<br /> * Other devices in the signal chain<br /> <br /> If the equipment changes state during the capture, the resulting model may be inaccurate or unusable.<br /> <br /> Tube amplifiers should be allowed to reach normal operating condition before critical captures are made.<br /> <br /> == Capturing pedals and preamps ==<br /> <br /> Pedals, preamps, and similar devices can often be returned directly to an audio interface input, provided that the electrical levels and connections are appropriate.<br /> <br /> A simple chain may therefore be:<br /> <br /> &#039;&#039;&#039;Interface output → reamp device → pedal or preamp → interface input&#039;&#039;&#039;<br /> <br /> Multiple devices may also be captured as a chain:<br /> <br /> &#039;&#039;&#039;Interface output → reamp device → pedal → preamp → processor → interface input&#039;&#039;&#039;<br /> <br /> In that case, the resulting model represents the combined behavior of the chain rather than each device individually.<br /> <br /> == Capturing an amplifier without a cabinet ==<br /> <br /> An amplifier output designed to drive a loudspeaker &#039;&#039;&#039;must not normally be connected directly to an audio interface input.&#039;&#039;&#039;<br /> <br /> A suitable load and signal-level conversion method is required unless the amplifier provides a dedicated output designed for this purpose.<br /> <br /> A direct amplifier capture might therefore use:<br /> <br /> &#039;&#039;&#039;Interface output → reamp/capture device → amplifier → suitable load/direct capture device → interface input&#039;&#039;&#039;<br /> <br /> This produces a model of the amplifier portion of the chain without necessarily including the acoustic behavior of a loudspeaker cabinet and microphone.<br /> <br /> A cabinet simulation or [[Cabinets and IRs|impulse response]] may then be used during playback.<br /> <br /> == Capturing an amplifier and cabinet ==<br /> <br /> To include a physical loudspeaker cabinet in the capture, the acoustic output of the cabinet must normally be converted back into an electrical signal with a microphone.<br /> <br /> A typical chain becomes:<br /> <br /> &#039;&#039;&#039;Interface output → reamp/capture device → amplifier → speaker cabinet → microphone → microphone preamp/interface input&#039;&#039;&#039;<br /> <br /> The resulting capture may contain the combined characteristics of:<br /> <br /> * Amplifier<br /> * Loudspeaker<br /> * Cabinet<br /> * Microphone<br /> * Microphone position<br /> * Microphone preamp<br /> * Other equipment in the return chain<br /> <br /> Moving the microphone or changing any of these components can substantially change the result.<br /> <br /> == Return path ==<br /> <br /> The returned signal must reach the capture system without unwanted clipping, processing, or alteration.<br /> <br /> Check for:<br /> <br /> * Correct interface input<br /> * Appropriate input type<br /> * Correct gain<br /> * Phantom power where appropriate<br /> * Unwanted high-pass filters<br /> * Automatic gain control<br /> * Software monitoring or processing<br /> * EQ or dynamics processing<br /> * Cabinet simulation<br /> * Sample-rate conversion<br /> <br /> Unless the capture procedure specifically requires processing, the safest general principle is to return the signal as directly and transparently as practical.<br /> <br /> == Gain staging ==<br /> <br /> Every stage in the chain has the potential to change level.<br /> <br /> A capture may involve several distinct level domains:<br /> <br /> # Digital level inside the computer<br /> # Interface output level<br /> # Signal presented to the device under test<br /> # Output level from the device under test<br /> # Interface input level<br /> # Recorded digital level<br /> <br /> These should not be treated as though they are interchangeable.<br /> <br /> For example, a recording that peaks safely below 0 dBFS tells you that the converter is not digitally clipping, but it does not tell you whether the amplifier or pedal received the intended input level.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Ground loops and noise ==<br /> <br /> Capture systems often connect a computer, audio interface, amplifier, and other mains-powered equipment simultaneously. This can create ground loops.<br /> <br /> Symptoms may include:<br /> <br /> * Hum<br /> * Buzz<br /> * Computer-related noise<br /> * High-frequency interference<br /> <br /> Balanced connections and transformer isolation can help where appropriate. A ground-lift facility on a properly designed reamp or capture device may also be useful.<br /> <br /> Do not defeat protective mains safety grounding as a solution to a ground-loop problem.<br /> <br /> == Monitoring ==<br /> <br /> Monitoring should be arranged so that it does not accidentally become part of the capture path.<br /> <br /> DAW routing can sometimes return the captured signal to an output, creating feedback or an unintended loop.<br /> <br /> Before beginning a capture, verify independently:<br /> <br /> * Which output carries the test signal<br /> * Which input receives the return<br /> * Where monitoring is routed<br /> * Whether any plugins are active<br /> * Whether direct monitoring is enabled<br /> <br /> == Document the chain ==<br /> <br /> A useful capture should be reproducible.<br /> <br /> When creating captures for yourself or others, record enough information to understand what the model represents.<br /> <br /> Useful documentation may include:<br /> <br /> * Device manufacturer and model<br /> * Device settings<br /> * Interface<br /> * Output used<br /> * Reamp or capture device<br /> * Input used<br /> * Calibration information<br /> * Cabinet and speaker<br /> * Microphone and position<br /> * Sample rate<br /> * Capture platform and version<br /> * Relevant software settings<br /> <br /> Good documentation can be almost as important as the model itself.<br /> <br /> == Troubleshooting the chain ==<br /> <br /> When a capture fails or sounds wrong, troubleshoot the chain one stage at a time.<br /> <br /> Do not immediately assume the modeling system failed.<br /> <br /> Verify:<br /> <br /> # The correct test signal is being played<br /> # The correct interface output is active<br /> # The signal reaches the device under test<br /> # The device is responding normally<br /> # The return signal reaches the intended interface input<br /> # Nothing clips<br /> # No unwanted processing is active<br /> # Levels and calibration meet the platform&#039;s requirements<br /> # The recorded or captured return has the expected duration and format<br /> <br /> Breaking the system into individual stages usually reveals problems much faster than repeatedly attempting the entire capture process.<br /> <br /> == See also ==<br /> <br /> * [[Getting Started with Capture Technology]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Troubleshooting Captures]]<br /> * [[Glossary]]<br /> <br /> &#039;&#039;&#039;[https://namforum.com/ Discuss capture signal chains on NAMFORUM]&#039;&#039;&#039;<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Capture techniques]]</div> Tue, 01 Sep 2026 05:47:14 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Capture_Signal_Chain https://namforum.com/wiki/index.php?title=Getting_Started_with_Capture_Technology&diff=3&oldid=0 https://namforum.com/wiki/index.php?title=Getting_Started_with_Capture_Technology&diff=3&oldid=0 <p>Created page</p> <p><b>New page</b></p><div>= Getting Started with Capture Technology =<br /> <br /> &#039;&#039;&#039;Capture technology&#039;&#039;&#039; is a method of creating a digital model of the behavior of real audio equipment or a signal chain.<br /> <br /> Instead of programming an amplifier, pedal, preamp, or other device component by component, a capture system measures how the equipment responds to known signals. The resulting data is used to create a model that can reproduce the behavior of the original equipment.<br /> <br /> Different systems use different terminology and different methods, but the basic idea is similar: &#039;&#039;&#039;send a known signal through something, measure what comes back, and create a model from the result.&#039;&#039;&#039;<br /> <br /> == What can be captured? ==<br /> <br /> Capture technology is most commonly associated with guitar amplifiers, but it is not inherently limited to guitar equipment.<br /> <br /> Depending on the capabilities of the particular system, possible capture targets can include:<br /> <br /> * Guitar and bass amplifiers<br /> * Amplifier and cabinet combinations<br /> * Preamps<br /> * Overdrive, distortion, boost, and other pedals<br /> * Studio equipment<br /> * Analog signal processors<br /> * Combinations of multiple devices<br /> * Complete signal chains<br /> <br /> As capture technology develops, its useful applications may extend well beyond the equipment for which it was originally designed.<br /> <br /> == Capture vs. traditional modeling ==<br /> <br /> Capture technology should not be confused with traditional algorithmic modeling.<br /> <br /> A traditional model attempts to reproduce a device using software designed around its known or measured behavior. The developer may model individual circuit stages, components, controls, nonlinearities, speaker behavior, or other characteristics.<br /> <br /> A capture system approaches the problem differently. It observes the response of a particular piece of equipment, usually at a particular configuration, and derives a model from those measurements.<br /> <br /> This distinction is important.<br /> <br /> A traditional model of an amplifier may provide functioning gain, tone, channel, and other controls intended to behave like the original amplifier.<br /> <br /> A capture, by contrast, commonly represents the behavior of the equipment at the settings used when the capture was created. Changing those settings significantly may therefore require another capture.<br /> <br /> The exact capabilities vary between platforms.<br /> <br /> == Major capture and profiling platforms ==<br /> <br /> Several technologies are widely used for capturing or profiling audio equipment.<br /> <br /> === Neural Amp Modeler (NAM) ===<br /> <br /> [[Neural Amp Modeler]] is an open-source project that uses machine learning to model nonlinear audio systems.<br /> <br /> NAM models can be created by recording the response of equipment to a supplied training signal and training a neural-network model from the resulting recording.<br /> <br /> See:<br /> <br /> * [[Neural Amp Modeler]]<br /> * [[Training a NAM Model]]<br /> * [[NAM Model File Format]]<br /> * [[NAM Playback]]<br /> <br /> === TONEX ===<br /> <br /> [[TONEX]] is IK Multimedia&#039;s capture and modeling ecosystem. Its Tone Models can represent amplifiers, cabinets, pedals, and combinations of equipment and can be used in TONEX software and compatible hardware.<br /> <br /> === Kemper PROFILER ===<br /> <br /> [[Kemper Profiling]] creates Profiles of amplifier and signal-chain configurations for use within the Kemper PROFILER ecosystem.<br /> <br /> === Neural Capture ===<br /> <br /> [[Neural Capture]] is Neural DSP&#039;s capture technology for creating digital representations of amplifiers, cabinets, and drive pedals for use with compatible Neural DSP hardware and software.<br /> <br /> === Line 6 Proxy ===<br /> <br /> [[Line 6 Proxy]] is Line 6&#039;s capture technology for creating models of amplifiers, pedals, and other equipment for compatible Line 6 systems.<br /> <br /> === Open capture technologies ===<br /> <br /> NAM is not the only open technology in this field. Other projects and research efforts are exploring machine-learning approaches to modeling audio equipment.<br /> <br /> See [[Open Capture Technologies]].<br /> <br /> == The basic capture signal chain ==<br /> <br /> Although platforms differ, a typical capture process looks something like this:<br /> <br /> &#039;&#039;&#039;Capture software or training signal → audio interface output → reamp/capture interface → device under test → audio interface input → recording or capture software&#039;&#039;&#039;<br /> <br /> For an amplifier being captured with its speaker cabinet, a microphone and microphone preamp may also be part of the return path.<br /> <br /> For a direct amplifier capture, the signal may instead be taken from a suitable load box or direct output.<br /> <br /> The details matter. Levels, routing, impedance, noise, clipping, sample rate, and the configuration of the equipment can all affect the resulting capture.<br /> <br /> See [[Capture Signal Chain]].<br /> <br /> == Calibration and levels ==<br /> <br /> One of the most important concepts in capture technology is &#039;&#039;&#039;level calibration&#039;&#039;&#039;.<br /> <br /> The level sent from an audio interface is not necessarily the same level that a guitar pickup would produce. Likewise, different interfaces, reamp devices, capture boxes, and input configurations can produce different electrical conditions.<br /> <br /> A capture can sound excellent while still being incorrectly calibrated.<br /> <br /> That becomes especially important when models are exchanged between users. If the capture process does not preserve meaningful input and output levels, the model may respond differently in somebody else&#039;s system.<br /> <br /> See [[Gain Staging and Calibration]].<br /> <br /> == Reamping ==<br /> <br /> Many capture workflows use techniques derived from traditional studio reamping.<br /> <br /> A recorded or generated signal leaves the audio interface and is sent into the equipment being captured. Depending on the equipment and interface, a reamp device may be used to provide appropriate level, connection, isolation, or grounding.<br /> <br /> The output of the device is then recorded back into the computer.<br /> <br /> See [[Reamping for Capture]].<br /> <br /> == Captures with and without cabinets ==<br /> <br /> Not every capture contains the same parts of the signal chain.<br /> <br /> An amplifier may be captured:<br /> <br /> * Directly, without a speaker cabinet<br /> * Through a cabinet and microphone<br /> * With a separate impulse response used for the cabinet<br /> * As part of a larger signal chain<br /> <br /> Understanding what is contained in a model is essential when choosing playback equipment and deciding whether an additional cabinet simulation or impulse response is required.<br /> <br /> See:<br /> <br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> <br /> == Creating your first capture ==<br /> <br /> Before attempting a first capture, understand the complete signal path.<br /> <br /> At minimum, identify:<br /> <br /> # What signal the capture system will generate or play<br /> # Which interface output will carry it<br /> # How that signal will reach the equipment being captured<br /> # How the equipment&#039;s output will be returned safely to the interface<br /> # Which input will record the result<br /> # What level and calibration requirements apply to the chosen capture system<br /> <br /> Do not assume that two connectors which physically fit together necessarily carry compatible signals.<br /> <br /> Once the routing is understood, the actual capture or training procedure is usually straightforward.<br /> <br /> See [[Creating a Capture]].<br /> <br /> == Evaluating a capture ==<br /> <br /> A successful capture is not simply one that completes without an error.<br /> <br /> Listen critically.<br /> <br /> Compare the model with the original equipment using the same source material and, where practical, matched playback conditions.<br /> <br /> Useful questions include:<br /> <br /> * Does the model have the same overall gain?<br /> * Does it respond similarly to changes in playing level?<br /> * Does the low end behave similarly?<br /> * Are high-frequency characteristics preserved?<br /> * Does distortion develop in the same way?<br /> * Are noise or artifacts present?<br /> * Are the input and output levels sensible?<br /> * Does the model remain convincing with different instruments or source material?<br /> <br /> A technically valid capture is not necessarily an accurate one.<br /> <br /> == Troubleshooting ==<br /> <br /> Capture problems often come from the signal chain rather than the modeling technology itself.<br /> <br /> Common causes include:<br /> <br /> * Incorrect routing<br /> * Incorrect input or output selection<br /> * Clipping<br /> * Signals that are too low<br /> * Incorrect calibration<br /> * Unwanted cabinet or speaker simulation<br /> * Ground loops<br /> * Excessive noise<br /> * Sample-rate or recording configuration problems<br /> * Equipment changing state during the capture<br /> * Incorrect trimming or processing of training recordings<br /> <br /> See [[Troubleshooting Captures]].<br /> <br /> == Learn, experiment, and share ==<br /> <br /> Capture technology is developing quickly. Documentation explains how a system is intended to work, but practical experience often reveals details that documentation does not.<br /> <br /> Experiment carefully. Document your signal chain and settings. Compare results. When something works particularly well—or fails in an interesting way—share what you learned.<br /> <br /> The &#039;&#039;&#039;[https://namforum.com/ NAMFORUM discussion forum]&#039;&#039;&#039; exists for exactly that purpose.<br /> <br /> The forum is where we discuss and experiment. The Wiki is where useful knowledge can become a lasting reference.<br /> <br /> == See also ==<br /> <br /> * [[Capture Technologies]]<br /> * [[Capture Signal Chain]]<br /> * [[Gain Staging and Calibration]]<br /> * [[Reamping for Capture]]<br /> * [[Creating a Capture]]<br /> * [[Capture Types]]<br /> * [[Cabinets and IRs]]<br /> * [[Glossary]]<br /> <br /> [[Category:Capture technology]]<br /> [[Category:Getting started]]</div> Tue, 01 Sep 2026 05:34:25 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Getting_Started_with_Capture_Technology https://namforum.com/wiki/index.php?title=Main_Page&diff=2&oldid=1 https://namforum.com/wiki/index.php?title=Main_Page&diff=2&oldid=1 <p>Main Page Creation</p> <table style="background-color: #fff; color: #202122;" data-mw-interface=""> <col class="diff-marker" /> <col class="diff-content" /> <col class="diff-marker" /> <col class="diff-content" /> <tr class="diff-title" lang="en"> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">← Older revision</td> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">Revision as of 01:31, 1 September 2026</td> </tr><tr><td colspan="2" class="diff-lineno" id="mw-diff-left-l1">Line 1:</td> <td colspan="2" class="diff-lineno">Line 1:</td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">&lt;strong&gt;MediaWiki has been installed.&lt;/strong&gt;</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">__NOTOC__</ins></div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><br></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><br></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">Consult the [https://www.mediawiki.org/wiki/Special:MyLanguage/Help:Contents User's Guide] for information on using the wiki software.</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">= Welcome to NAMFORUM Wiki =</ins></div></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><br></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><br></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>== Getting <del style="font-weight: bold; text-decoration: none;">started </del>==</div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">The '''NAMFORUM Wiki''' is a community knowledge base for '''Neural Amp Modeling technology and other capture technologies'''.</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>* [<del style="font-weight: bold; text-decoration: none;">https://www</del>.<del style="font-weight: bold; text-decoration: none;">mediawiki</del>.<del style="font-weight: bold; text-decoration: none;">org/wiki/Special:MyLanguage/Manual:Configuration_settings Configuration settings list]</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">* </del>[https://<del style="font-weight: bold; text-decoration: none;">www.mediawiki</del>.<del style="font-weight: bold; text-decoration: none;">org/wiki/Special:MyLanguage</del>/<del style="font-weight: bold; text-decoration: none;">Manual:FAQ MediaWiki FAQ</del>]</div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">Capture and profiling technologies make it possible to reproduce the behavior of amplifiers, pedals, preamps, cabinets, studio equipment, complete signal chains, and other audio systems. This Wiki documents the technologies, equipment, techniques, terminology, and practical knowledge behind creating and using those captures.</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>* [<del style="font-weight: bold; text-decoration: none;">https</del>:<del style="font-weight: bold; text-decoration: none;">//lists.wikimedia.org/postorius/lists/mediawiki-announce.lists.wikimedia.org/ MediaWiki release mailing list</del>]</div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>* [<del style="font-weight: bold; text-decoration: none;">https</del>:<del style="font-weight: bold; text-decoration: none;">//www.mediawiki.org/wiki/Special:MyLanguage/Localisation#Translation_resources Localise MediaWiki for your language</del>]</div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">'''New to capture technology?''' Start with [[Getting Started with Capture Technology]].</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>* [<del style="font-weight: bold; text-decoration: none;">https</del>:<del style="font-weight: bold; text-decoration: none;">//www</del>.<del style="font-weight: bold; text-decoration: none;">mediawiki</del>.<del style="font-weight: bold; text-decoration: none;">org/wiki/Special:MyLanguage/Manual:Combating_spam Learn how to combat spam on your wiki]</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div>== <ins style="font-weight: bold; text-decoration: none;">Capture Fundamentals ==</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[</ins>Getting <ins style="font-weight: bold; text-decoration: none;">Started with Capture Technology]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Capture Signal Chain]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Gain Staging and Calibration]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Reamping for Capture]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Creating a Capture]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Capture Types]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Cabinets and IRs]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Troubleshooting Captures]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">== Capture Platforms ==</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">=== Neural Amp Modeler ===</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Neural Amp Modeler]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[NAM Model File Format]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Training a NAM Model]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[NAM A2 Architecture]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[NAM Playback]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">=== Commercial Capture Systems ===</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[TONEX]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Kemper Profiling]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Neural Capture]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Line 6 Proxy]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">=</ins>== <ins style="font-weight: bold; text-decoration: none;">Open Capture Technologies ===</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Open Capture Technologies]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">== Reference ==</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Capture Technologies]] — overview and comparison of capture and profiling technologies</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">* [[Glossary]] — terminology used throughout the Wiki</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div>* [<ins style="font-weight: bold; text-decoration: none;">[Resources]] — software, documentation, model libraries, communities, and other useful links</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">== NAMFORUM Community ==</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">The Wiki and the NAMFORUM discussion forum serve different but complementary purposes.</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">The '''forum''' is where we discuss, experiment, ask questions, share experiences, and make discoveries</ins>.</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">The '''Wiki''' is where established information and useful knowledge can be organized into a lasting reference</ins>.</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">'''</ins>[https://<ins style="font-weight: bold; text-decoration: none;">namforum</ins>.<ins style="font-weight: bold; text-decoration: none;">com</ins>/ <ins style="font-weight: bold; text-decoration: none;">Visit the NAMFORUM discussion forum</ins>]<ins style="font-weight: bold; text-decoration: none;">'''</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">Useful discussions from the forum may be referenced throughout the Wiki so readers can examine the conversations and experiences behind the information.</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">== About This Wiki ==</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div>* [<ins style="font-weight: bold; text-decoration: none;">[Meta</ins>:<ins style="font-weight: bold; text-decoration: none;">Contributors|Contributors]</ins>]</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div>* [<ins style="font-weight: bold; text-decoration: none;">[Meta</ins>:<ins style="font-weight: bold; text-decoration: none;">Administrators|Administrators]</ins>]</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div>* [<ins style="font-weight: bold; text-decoration: none;">[Meta</ins>:<ins style="font-weight: bold; text-decoration: none;">History|History of NAMFORUM Wiki]]</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">== Help Build the Wiki ==</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">Capture technology is developing rapidly, and much of the most useful knowledge comes from people actually using it.</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">If you have experience with NAM, TONEX, Kemper, Neural Capture, Line 6 Proxy, open-source capture systems, or related technologies, your knowledge can help make this resource better.</ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div> </div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">'''Share what you know</ins>. <ins style="font-weight: bold; text-decoration: none;">Document what you learn</ins>.<ins style="font-weight: bold; text-decoration: none;">'''</ins></div></td></tr> </table> Tue, 01 Sep 2026 05:31:55 GMT NAMFORUM Sysop https://namforum.com/wiki/index.php/Talk:Main_Page