diff --git a/README.md b/README.md index ee3baa3..e12bd8a 100644 --- a/README.md +++ b/README.md @@ -13,7 +13,7 @@ Download: v2.0. Website: [https://rerdavies.github.io/pipedal](https://rerdavies.github.io/pipedal). Documentation: [https://rerdavies.github.io/pipedal/Documentation.html](https://rerdavies.github.io/pipedal/Documentation.html). -#### Announcing PiPedal 2.0 (2.0.103)—a major update to PiPedal, including big new features. See the [documentation](https://rerdavies.github.io/pipedal/PiPedal.html) for information, and instructions on how to install Pipedal 2.0. +#### Announcing PiPedal 2.0 (2.0.103)—a major update to PiPedal, including exciting new features, ioncluding support for NAM A2. See the Pipedal website [documentation](https://rerdavies.github.io/pipedal/PiPedal.html) for more information.   diff --git a/artifacts/model_comparison.png b/artifacts/model_comparison.png new file mode 100644 index 0000000..44dca7d Binary files /dev/null and b/artifacts/model_comparison.png differ diff --git a/artifacts/model_comparison.svgx b/artifacts/model_comparison.svgx new file mode 100644 index 0000000..d53b199 --- /dev/null +++ b/artifacts/model_comparison.svgx @@ -0,0 +1,200 @@ + + + + + + + + + + + + + + + + + + + + + + + Line 6 Proxy + 58.55 + IK Multimedia TONEX + 66.82 + Neural DSP Neural Capture V2 + 74.69 + NAM A2 Lite + 77.18 + NAM A1 Standard + 78.37 + NAM A2 Full + 81.82 + diff --git a/artifacts/model_comparison.svgx.png b/artifacts/model_comparison.svgx.png new file mode 100644 index 0000000..9313c92 Binary files /dev/null and b/artifacts/model_comparison.svgx.png differ diff --git a/docs/ReleaseNotes.md b/docs/ReleaseNotes.md index 6c8f0ee..a849b1e 100644 --- a/docs/ReleaseNotes.md +++ b/docs/ReleaseNotes.md @@ -4,10 +4,13 @@ PiPedal 2.0 is a big release, with exciting new features, improvements, and bug fixes. -The headline feature of PiPedal 2.0 is support for TooB Neural Amp Modeler A2 models. TooB Neural Amp Modeler A2 models provide significant improvements in sound quality and performance compared to NAM A1 models, while consuming significantly less CPU resources. This is particularly important for users of Raspberry Pi 4, since NAM A2 models can provide better sound quality than NAM A1 models, while consuming significantly less CPU resources. +The headline feature of PiPedal 2.0 is support for TooB Neural Amp Modeler A2 models. TooB Neural Amp Modeler A2 models provide significant improvements in sound quality and performance compared to NAM A1 models, while consuming significantly less CPU resources. This is particularly important for users of Raspberry Pi 4, since it allows Pi 4 users to use an additional NAM instance in a preset. -A new integrated Tone3000 model downloader provides a convenient way to download the latest NAM A2 models and IIR files from Tone3000 directly to the PiPedal server. All existing NAM A1 models on Tone3000 have been retrained for NAM A2. When -you download from Tone3000 using the integrated downloader, you will get NAM A2 models. Just click on the DOWNLOAD MODELS FROM TONE3000 button. + +PiPedal 2.0 connects to [Tone3000.com](https://www.tone3000.com) web services, so that you can +download Tone3000 NAM models and IRs directly to the PiPedal server, without ever having to leave +the PiPedal web application interface. All existing NAM A1 models on Tone3000 have been retrained for NAM A2. When you download from Tone3000 using the integrated downloader, you will get NAM A2 models. Just click on the DOWNLOAD MODELS FROM TONE3000 button in the File Selection dialog for Toob Neural Amp Modeler, in order to download NAM A2 models directly +to the PiPedal server. Currently, the best way to get access to NAM A2 models is via the integrated Tone3000 model downloader. Models downloaded externally from the Tone3000 website will still be NAM A1 models; however this should change soon with a forthcoming announcement from Tone3000. @@ -33,7 +36,7 @@ PiPedal 2.0 can now be installed as a Progressive Web Application (PWA). This al - Copy and Paste plugins (and their settings) between pedalboards. Accessible via the "Copy Plugin" and "Paste Plugin" menu items in the former Plugin Preset selector menu. -### Minor features and bug fixes +### Miscellaneous features and bug fixes - Runtime detection and use of AVX extensions for TooB plugins. Reduces CPU use of TooB NAM by 50% on most x86_64/amd64 computers. diff --git a/docs/img/model_comparison.png b/docs/img/model_comparison.png new file mode 100644 index 0000000..9313c92 Binary files /dev/null and b/docs/img/model_comparison.png differ diff --git a/docs/index.md b/docs/index.md index b57e7de..808e4e9 100644 --- a/docs/index.md +++ b/docs/index.md @@ -21,7 +21,7 @@ instructions](https://rerdavies.github.io/pipedal/Configuring.html) to make sure {% include gallery.html %} -New in PiPedal v2.0: +## New in PiPedal v2.0: - Support for Neural Amp Modeler (NAM) A2 models. - Direct single-step downloads of NAM A2 models to the Pipedal server using web services provided by Tone3000.com. @@ -29,11 +29,22 @@ New in PiPedal v2.0: - A new Channel Routing dialog which allows you to pass through Auxilliary audio channels, or unprocessed guitar inputs for later re-amping in a DAW or external hardware. - New NAM A2-based Factory Presets, and a small selection of NAM A2 models pre-installed and ready to use. -PiPedal includes state-of-the-art AI-based guitar amp emulation, using the TooB Neural Amp Modeler technology. And PiPedal 2.0 now includes support for the brand new NAM A2 technology, which provides event more accurate amp simulations than NAM A1, while using even less CPU. Experience the ground-breaking quality of NAM A2 models now, with PiPedal's low-latency audio engine running on your Raspberry Pi or Ubuntu computer. +PiPedal includes state-of-the-art AI-based guitar amp emulation, using the TooB Neural Amp Modeler plugin, which now includes support for the brand new NAM A2 technology. NAM A2 provides event more accurate amp simulations than NAM A1, while using even less CPU. Experience the ground-breaking quality of NAM A2 models now, with PiPedal's low-latency audio engine running on your Raspberry Pi or Ubuntu AMD64/x86_64 computer. -NAM changes everything! The quality of NAM A1 and A2 models is better than than amp emulations on top-of-the-line commercial guitar stomp boxes costing thousands of dollars. Simulations that not only sound like the real thing, but also respond to your playing dynamics in the same way as the real amp. +NAM changes everything! Again! NAM A2 models provide even better quality, and even lower CPU use that the previous NAM A1 models did. And NAM A2 provides amp and effect emulations that are demonstrably, and measurably better than competing commercial amp modeling technologies. And it is available for free to PiPedal users now! -PiPedal 2.0 integrates with Tone3000.com's web services, allowing you to directly install new NAM A2 models on the pipedal server without ever leaving the PiPedal user interface. Or download and install commercially-developed NAM models from a rich ecosystem of model providers. + + +

+ Fig 1: Unscreened ratings from a large-scale blind MUSHRA listening test evaluating + NAM A2 amp/effect modeling against other commercial + modelers. 105,842 ratings from + 1,184 participants across 37 tones. Data provided by TONE3000 and Steve Atkinson under a CC-BY 4.0 license. + 1 +

+ + +PiPedal 2.0 integrates with Tone3000.com's web services, allowing you to directly install new NAM A2 models on the pipedal server without ever leaving the PiPedal user interface. Or you can download and install commercially-developed NAM models from a rich ecosystem of model providers. {% include demo.html %} @@ -49,6 +60,20 @@ PiPedal includes a pre-installed selection of LV2 plugins from the ToobAmp colle If your USB audio adapter has MIDI connectors, you can use MIDI devices (keyboards, controllers, or midi floor boards) to control PiPedal while performing. A simple interface allows you to select how you would like to bind PiPedal controls to midi messages. +
+

+ 1 TONE3000, & Atkinson, S. (2026). A2 MUSHRA Listening + Test Raw Data [Data set]. + Tone3000. + https://www.tone3000.com/guides/nam-a2-the-complete-guide + + . Repository: + + https://github.com/tone-3000/a2-mushra-data + + . License: CC BY 4.0. ↩ +

+