Neural Amp Modeler NAM amp capture/profile player FREE *December 14, 2023 update 0.7.7

[jamcat]

I should also add, that if you have amps that you can capture with a re-amp box with the provided training file but don’t have the ability to convolve them in the training software to convert them, go onto the NAM FB group and kindly ask one of the guys there that have been doing full captures/training conversions, and they could process your files for you. It’s all about getting more and unique and diverse amps into this free community database of profiles. It’s growing with more new users and even more profile packs each day.

I’m a bit confused by the use of “convolve” in the context of machine learning, as there is no convolution (f * g) involved in training. It’s just incrementally adding or subtracting weights depending on if it’s getting closer or further from the training data until it converges on a model that can’t be improved upon.

Are you seeing other people calling this convolution, or is that just your term?

[metalifuxx]

I should also add, that if you have amps that you can capture with a re-amp box with the provided training file but don’t have the ability to convolve them in the training software to convert them, go onto the NAM FB group and kindly ask one of the guys there that have been doing full captures/training conversions, and they could process your files for you. It’s all about getting more and unique and diverse amps into this free community database of profiles. It’s growing with more new users and even more profile packs each day.

I’m a bit confused by the use of “convolve” in the context of machine learning, as there is no convolution (f * g) involved in training. It’s just incrementally adding or subtracting weights depending on if it’s getting closer or further from the training data until it converges on a model that can’t be improved upon.

Are you seeing other people calling this convolution, or is that just your term?

brain fart, i meant "process" not "convolve". Someone in the NAM group was talking about shooting and convolving cab IRs at the same time, so that was a Freudian slip, doh

BTW, i am just an ambassador/messenger for the NAM group, not a dev in anyway

[zerocrossing]

Well, color me curious. I’m no amp purest, but I sure do love my AxeFX 3. Not sure I’d want to ditch it but I’m always up for something new. I tried the Neural plugins during the gap between my Kemper and the AxeFX (I actually really liked the Kemper with the right profile, but wanted to go back to my old dual amp setup) and found them pretty good.

[ohdoubleu]

Ah, this is awesome! I'm super impressed by all the HM-2-style pedals and the AC15/30 amps, I've always found these especially tricky to nail in software. Huge thanks to Steven Atkinson - if he ever checks out this thread - and an extra special thanks to metalifuxx for taking the time to put together the profile zips, they're much appreciated! One small question, I can see the Clean Deluxe Reverb in the collection but was the Fuzz into Deluxe Reverb mentioned on the first page ever ported to the new all-in-one format? Thanks!

[metalifuxx]

One small question, I can see the Clean Deluxe Reverb in the collection but was the Fuzz into Deluxe Reverb mentioned on the first page ever ported to the new all-in-one format? Thanks!

Good question. That I’m not sure of, possibly it was converted to the new single format and i missed it in one of the model posts. Somewhere in the group or on the Steve Ack github, there is a server utility that allows you to convert the formats by uploading the *.json file and the weights file and it will spit out a *.nam single file. That may still have to be done. He did port that as an individual vst3 file originally though, so there is still that if those links are still active.

[ohdoubleu]

One small question, I can see the Clean Deluxe Reverb in the collection but was the Fuzz into Deluxe Reverb mentioned on the first page ever ported to the new all-in-one format? Thanks!

Good question. That I’m not sure of, possibly it was converted to the new single format and i missed it in one of the model posts. Somewhere in the group or on the Steve Ack github, there is a server utility that allows you to convert the formats by uploading the *.json file and the weights file and it will spit out a *.nam single file. That may still have to be done. He did port that as an individual vst3 file originally though, so there is still that if those links are still active.

Ah, no worries! There's plenty in the pack to keep me busy but I'll definitely check out the standalone vst3s if I'm still pining for fuzzes... I just really love fuzzes! And, based on the Big Muff in the collection, I'm sure that NAM would do them all great justice. Thanks for your help though!

[raintalk]

This is really quite impressive. It sounds spot on.

I'm curious if after the capture is done, can a lower powered PC be used for playback mode? There are many low cost mini PCs around with enough power to run DAWs and audio. It would be interesting to know if one could be easy enough to use in live situations.

[metalifuxx]

This is really quite impressive. It sounds spot on.

I'm curious if after the capture is done, can a lower powered PC be used for playback mode? There are many low cost mini PCs around with enough power to run DAWs and audio. It would be interesting to know if one could be easy enough to use in live situations.

I have not tested it yet on older systems. I have a few older laptops circa 2013 laying around that i have setup for mobile audio and plugins, and also a 2015 Macbook.

[zerocrossing]

I tried a couple of IR files I got from this site:

https://www.forward-audio.com/free-impu ... dern-rock/

and I get an error telling me "wave file contents are invalid." I opened it up in Adobe Audition and it seemed fine. Not sure what I'm doing wrong. Could someone post a place I could grab a working IR to test it?

[metalifuxx]

I tried a couple of IR files I got from this site:

https://www.forward-audio.com/free-impu ... dern-rock/

and I get an error telling me "wave file contents are invalid." I opened it up in Adobe Audition and it seemed fine. Not sure what I'm doing wrong. Could someone post a place I could grab a working IR to test it?

The IR loader is still buggy, or more so particular with the right type/format of IR, I forget exactly what format/length they have to be. Best to use an IR loader plugin after NAM for right now.

[zerocrossing]

I tried a couple of IR files I got from this site:

https://www.forward-audio.com/free-impu ... dern-rock/

and I get an error telling me "wave file contents are invalid." I opened it up in Adobe Audition and it seemed fine. Not sure what I'm doing wrong. Could someone post a place I could grab a working IR to test it?

The IR loader is still buggy, or more so particular with the right type/format of IR, I forget exactly what format/length they have to be. Best to use an IR loader plugin after NAM for right now.

Interesting. I thought it had to have an IR loaded to process the sound, because it absolutely does nothing on my PC. I did find an IR that worked (oddly one for an Oculus Quest that I use to preview my VR mixes) but nothing in my sound changed at all. Bitwig, latest build on Windows 10, latest build.

[metalifuxx]

The GUI trainer is here

Folks who are training on their own computers can get the latest here: https://github.com/.../neural-amp.../re ... tag/v0.5.0
After following the install steps, if you run "nam" in the terminal, you'll be greeted by this lovely GUI!

[metalifuxx]

edit

Get free profiles now from

www.tonehunt.org

[naturestoned]

what a great project, thanks for sharing and the efforts from all who contributed #thumbsup

quality is impressing. had a quick tryout with the very nice JCM800 v1.1 #1 at the weekend together with a York Audio MV30-Dual IR.

I have a Friedman Dirty Shirley, Suhr RL and some other nice tube stuff and I'd like to try to capture it myself if I find some time. Still need a reamp box though, having an eye at the Palmer daccapo

do the captures automatically translate to all sample rates or would it be best to play them at capture sample rate for full accuracy?

[eli999]

I have troubles getting the standalone app to receive input from my audio interface. Tried switching inputs, creating aggregate device etc., the in-built laptop mic works, but the interface inputs don't in any configuration. Audient iD4 interface, macOS Big Sur. The plugin in a daw does work as expected though. Any ideas?