TONE3000 - tone3000.com (previously ToneHunt and TONEZONE3000)

[deadpool_25]

I went on there and grabbed a few tones to try with my GP-5. Easy to use even though it was my first time on the site. I didn't even look at any of the guides. Just downloaded a few tones and was off and running. I'll check those guides out though.

Edit: or not lol

Edit 2: that was the get-started link. The others work. And the "Get Started" link on the 404 page works.

Edit 3: the correct link looks to be https://www.tone3000.com/guides/get-started-with-tone3000

 

[staas]

I went on there and grabbed a few tones to try with my GP-5. Easy to use even though it was my first time on the site. I didn't even look at any of the guides. Just downloaded a few tones and was off and running. I'll check those guides out though.

Edit: or not lol
View attachment 53558

Edit 2: that was the get-started link. The others work. And the "Get Started" link on the 404 page works.

Edit 3: the correct link looks to be https://www.tone3000.com/guides/get-started-with-tone3000

Sorry about that! Thanks for catching. I updated the previous message.

 

[staas]

We’re launching a beginner-friendly guide that shows how to capture Neural Amp Modeler profiles of your amps, pedals, outboard gear, and signal chains using the sweep signal method.

Next up, we’ll release guides covering the dry/wet method.

Guide: https://www.tone3000.com/guides/capture-your-gear-with-nam-sweep-method

Video:

 

[staas]

Now we're launching a beginner-friendly guide that shows how to capture Neural Amp Modeler profiles of your amps, pedals, outboard gear, and signal chains using the Dry/Wet method. Unlike the Sweep Signal method, you don't need a reamp box: just something to split your signal (like a tuner pedal) and an interface with two inputs.

You can also use this method with old sessions! If you have previously recorded dry and wet stems from an old session, you can upload those instead of recording new ones. I've personally done this from sessions that are 10+ years old!

Guide: https://www.tone3000.com/guides/capture-your-gear-with-nam-dry-wet

 

[Richard_G]

Hmm… I’d guess old sessions wouldn’t work as well as specialized training files—unless those sessions have some really weird-ass guitar parts!

 

[Sascha Franck]

Hmm… I’d guess old sessions wouldn’t work as well as specialized training files—unless those sessions have some really weird-ass guitar parts!

Likely.
I'll be giving it a try with some "standard" guitar parts later on, though - and then check what the captures will be like in case I'm playing something pretty different and use a different guitar.

 

[staas]

Hmm… I’d guess old sessions wouldn’t work as well as specialized training files—unless those sessions have some really weird-ass guitar parts!

They work just fine, especially if you're going to be playing similar parts. Try it!

 

[staas]

TONE3000 and Steve Atkinson are partnering to launch the next generation of Neural Amp Modeler. It's called A2 and it will be a more efficient architecture that will run on more devices and sound even better. The key innovation: no more standard/feather/nano versions. NAM files now automatically scale to match your device's processing capabilities. One NAM captures adapts to any hardware, from high-powered computers to resource-constrained pedals.

https://www.tone3000.com/blog/a2-next-generation-neural-amp-modeler

 

[staas]

Thrilled to announce that Blackstar has unveiled the Beam Mini – the world's first amplifier to natively support Neural Amp Modeler (NAM) via direct integration with TONE3000. The Beam Mini will support Architecture 2, the next generation of Neural Amp Modeler.

https://www.tone3000.com/blog/blackstar-unveils-beam-mini-powered-by-tone3000

 

[Sascha Franck]

Thrilled to announce that Blackstar has unveiled the Beam Mini – the world's first amplifier to natively support Neural Amp Modeler (NAM) via direct integration with TONE3000. The Beam Mini will support Architecture 2, the next generation of Neural Amp Modeler.

https://www.tone3000.com/blog/blackstar-unveils-beam-mini-powered-by-tone3000

Oh wow! Who would've thought that Blackstar would lead that race.

Now give me a small, pedalboardfriendly pedal along the lines of a Tonex One.

 

[Elric]

So I wonder if Blackstar is bankrolling this 'free' retraining to A2 effort? Looks like this is using the A2 stuff because it is more efficient on the player side.

Kind of interesting seeing something so 'low end' supporting NAM directly when existing high end hardware cannot support A1. I wonder if the A2 stuff would open the door for @FractalAudio to support it on AxeFx III (or other units like Stadium) to run this flavor of profile in block. They almost certainly have more DSP than this thing.

 

[Chocol8]

Looks like this is using the A2 stuff because it is more efficient on the player side.

How much it is due to A2 being more efficient vs A2 allowing the player to scale down quality without needing a conversion like the existing solutions is yet to be seen, but I would be shocked if this device is running A2 native at max quality.

 

[ian_dissonance]

That checks out after messing having the Beam Solo for the last week. I’m guessing (GUESSING) that the “Ampton” models in there of non-Blackstar amps are some sort of capture. The controls are not super functional, similar to my experience with NAM. Their 5150 model has basically the exact same amount of gain at 2 as it does at 10.

 

[staas]

Check out this new blog post from TONE3000’s João Felipe Santos on running Neural Amp Modeler (NAM) on an embedded device (Electrosmith Daisy Seed): https://www.tone3000.com/blog/running-nam-on-embedded-hardware

Along the way, performance improvements were contributed back to NAM Core, and several tools were open-sourced so others can build their own ARM Cortex neural NAM pedal.

NAM Pedal (Daisy Seed guide)

GitHub - tone-3000/nam-pedal: A demo of NeuralAmpModeler running on a Daisy Seed board

A demo of NeuralAmpModeler running on a Daisy Seed board - tone-3000/nam-pedal

github.com

NAM Binary Loader (compact model format)

GitHub - tone-3000/nam-binary-loader: Binary file format and loader for Neural Amp Modeler

Binary file format and loader for Neural Amp Modeler - tone-3000/nam-binary-loader

github.com

João’s technical deep dive (benchmarks + learnings)

João Felipe Santos | Bringing NAM A2 to Embedded Hardware

Contributions to NAM Core

Add inline GEMM optimizations and general performance improvements (#… · sdatkinson/NeuralAmpModelerCore@a0c93c0

…226) * Add inline GEMM optimizations and general performance improvements Hand-optimized GEMM kernels for small matrices common in NAM models, gated by #ifdef NAM_USE_INLINE_GEMM with Eigen fall...

github.com

This work supports the development of Steve Atkinson’s next-generation NAM Architecture 2 (A2).

Announcing A2: The Next Generation of Neural Amp Modeler · TONE3000

TONE3000 is partnering with the creator of Neural Amp Modeler, Steve Atkinson, to build the future of NAM so it's more efficient, works on more devices, and sounds even better. Launching March 2026.

www.tone3000.com

 

[2dor]

Check out this new blog post from TONE3000’s João Felipe Santos on running Neural Amp Modeler (NAM) on an embedded device (Electrosmith Daisy Seed):

https://www.tone3000.com/blog/running-nam-on-embedded-hardware

---

Open-source tools released for the community​

Along the way, performance improvements were contributed back to NAM Core, and several tools were open-sourced so others can build their own ARM Cortex neural NAM pedal.

NAM Pedal (Daisy Seed guide)
https://github.com/tone-3000/nam-pedal

NAM Binary Loader (compact model format)
https://github.com/tone-3000/nam-binary-loader

João’s technical deep dive (benchmarks + learnings)
https://jfsantos.dev/blog/nam-a2-embedded

Contributions to NAM Core
https://github.com/sdatkinson/NeuralAmpModelerCore/commit/a0c93c0c3934abec0bca80d2e9e592ac86217db4

---

Part of the Architecture 2 (A2) initiative​

This work supports the development of Steve Atkinson’s next-generation NAM Architecture 2 (A2).

Learn more about A2:
https://www.tone3000.com/blog/a2-next-generation-neural-amp-modeler

Massive drop! Should open a LOT of doors for small builders.

 

[AlbertA]

Massive drop! Should open a LOT of doors for small builders.

That's nice - I was starting to look into Eigen - but looks like they are on top of that already, very nice.

 

[staas]

TONE3000 just launched Live Input, a new feature that lets you plug your guitar or bass into your audio interface and play through any of the 275,000 NAM captures and IR's directly in your web browser.

Introducing Live Input on TONE3000

Plug in your guitar or bass and play through 275K tones directly in your browser. Now in beta.

www.tone3000.com

No need to download files or load them into a DAW first. Just connect your instrument, hit “Connect” on a tone page, and hear how it responds to your playing in real time. The same Neural Amp Modeler engine used in plugins and hardware powers this feature, and we released the underlying WebAssembly code open source if anyone wants to build on it.

Live Input is in beta so please try it and let us know what you think!

 

[paisleywookiee]

TONE3000 just launched Live Input, a new feature that lets you plug your guitar or bass into your audio interface and play through any of the 275,000 NAM captures and IR's directly in your web browser.

Introducing Live Input on TONE3000

Plug in your guitar or bass and play through 275K tones directly in your browser. Now in beta.

www.tone3000.com

No need to download files or load them into a DAW first. Just connect your instrument, hit “Connect” on a tone page, and hear how it responds to your playing in real time. The same Neural Amp Modeler engine used in plugins and hardware powers this feature, and we released the underlying WebAssembly code open source if anyone wants to build on it.

Live Input is in beta so please try it and let us know what you think!

Well, I can preemptively say I got nothing done tomorrow.

 

[dws2384]

TONE3000 just launched Live Input, a new feature that lets you plug your guitar or bass into your audio interface and play through any of the 275,000 NAM captures and IR's directly in your web browser.

Introducing Live Input on TONE3000

Plug in your guitar or bass and play through 275K tones directly in your browser. Now in beta.

www.tone3000.com

No need to download files or load them into a DAW first. Just connect your instrument, hit “Connect” on a tone page, and hear how it responds to your playing in real time. The same Neural Amp Modeler engine used in plugins and hardware powers this feature, and we released the underlying WebAssembly code open source if anyone wants to build on it.

Live Input is in beta so please try it and let us know what you think!

Wow

 

[AlbertA]

TONE3000 just launched Live Input, a new feature that lets you plug your guitar or bass into your audio interface and play through any of the 275,000 NAM captures and IR's directly in your web browser.

Introducing Live Input on TONE3000

Plug in your guitar or bass and play through 275K tones directly in your browser. Now in beta.

www.tone3000.com

No need to download files or load them into a DAW first. Just connect your instrument, hit “Connect” on a tone page, and hear how it responds to your playing in real time. The same Neural Amp Modeler engine used in plugins and hardware powers this feature, and we released the underlying WebAssembly code open source if anyone wants to build on it.

Live Input is in beta so please try it and let us know what you think!

Pretty cool - worked on Safari for me - but didn't on Firefox 148 both running on macOS Tahoe 26.3
Plenty of latency when I tried it on Safari.