NAM: Neural Amp Modeler

Also the extent to which aliasing is exhaustively discussed would make you think it was far more of an issue than it actually is. You'd think it was the bane of digital modeling when in fact these days its almost never making a significant enough difference for most people to notice or recognize as aliasing at all.

This is true. At least in the test I did, the difference was subtle enough that if I didn't have the actual amp to test with I would probably not notice. And how much of the difference in fidelity was caused by a reduction in aliasing, and how much was caused by a reduction in ESR I don't know yet.

One day I _will_ have a perfect model of the D. B. Cooper amp!

With modern gear if you want to achieve / get lower aliasing - I've got no issue with that :)

But I could list a bucket load of things that are going to bug most people more than aliasing levels in 2025 gear - some that quickly come to mind - how noisy are your pickups - how long a guitar lead do you use - going wireless - what kind of pick do you play with - how do you attack the strings - what is the quality of the analog front and back end of your modeler - how do you amplify and hear your modeler etc.....
 
If people want to modify NAM and make training 5 x longer and files sizes 10 x bigger to get %2 - %3 better aliasing - good luck to them :)

NAM is already - compared to Tonex - a CPU hog - imagine how few people would even bother trying to run it at those levels ?

And then, when is enough enough - people will then no doubt want to make models that take 10 x longer to train and be 30 x bigger in file size to get another %1 better aliasing ?
That's not how it works though, training with those super inputs doesn't increase cpu usage at all, training with super inputs plus xSTD architecture only increases cpu usage by ~10% compared to standard architecture, but that happens even with standard input on that architecture.
So it's an almost "free" upgrade on nam, cpu-wise.

If you and most people don't care about aliasing, go ask line 6 and cliff to remove oversampling from their products, an axe fx could probably run 32 amp blocks in a preset with no oversampling.
 
Last edited:
That's not how it works though, training with those super inputs doesn't increase cpu usage at all, training with super inputs plus xSTD architecture only increases cpu usage by ~10% compared to standard architecture, but that happens even with standard input on that architecture.
So it's an almost "free" upgrade on nam, cpu-wise.

If you and most people don't care about aliasing, go ask line 6 and cliff to remove oversampling from their products, an axe fx could probably run 32 amp blocks in a preset with no oversampling.

Hey. Not saying that at all.

I'm just of the view that the way Fractal have it -and L6 since FW 3.1 - it is at such a low level in 2025 that there are numerous other things that impact on the tone way more than any aliasing that occurs in the -40db <-> -70db range.

And in short, it seems like that is what Steve A. was getting at - amongst other things - in his blog post.

But regardless, I have less than zero issue with anyone who wants to make NAM Captures better for themselves or others for whatever reason.

It just reads to me that Steve A. does not see this as a priority vs pushing broader NAM adoption.

And as I've said before, for > 2 Months I've been running Amalgam NAM Captures in Gigfast Lite via AUM Mixer on my IPad Air M1 at 48k / 32 Samples with an IRig Pro Duo and MC6 Midi FSW - totally rock solid stable - sounds great and the aliasing that I know definitely exists - is totally and utterly un-hearable

I A/B'd 2 of Slammin Mofo's Standard and Complex Captures - running Amp only I could hear a very slight difference - true - but the second I put it into a full-patch with pre and post Amp Blocks ie: EQ's, Drive's, Comps. Delays, Rev's, Mods etc.... you would have needed "Jesus Ears" to tell the difference ie: there wasn't any even though I know/knew it was definitely there.

My apologies if my post mis-represented what I meant to "say" :)
 
Hey. Not saying that at all.
Well, it looked like it :D
I'm just of the view that the way Fractal have it -and L6 since FW 3.1 - it is at such a low level in 2025 that there are numerous other things that impact on the tone way more than any aliasing that occurs in the -40db <-> -70db range.
Totally agree here
And as I've said before, for > 2 Months I've been running Amalgam NAM Captures in Gigfast Lite via AUM Mixer on my IPad Air M1 at 48k / 32 Samples with an IRig Pro Duo and MC6 Midi FSW - totally rock solid stable - sounds great and the aliasing that I know definitely exists - is totally and utterly un-hearable

I A/B'd 2 of Slammin Mofo's Standard and Complex Captures - running Amp only I could hear a very slight difference - true - but the second I put it into a full-patch with pre and post Amp Blocks ie: EQ's, Drive's, Comps. Delays, Rev's, Mods etc.... you would have needed "Jesus Ears" to tell the difference ie: there wasn't any even though I know/knew it was definitely there.
Well, you're the one who couldn't hear the high frequency roll-off on tonex, so... I'll leave it at that :p
 
I’m looking at NAM XT for use with an HX Stomp. Could someone confirm if it’s possible to set whichever input and output on your audio interface in the standalone app on iPhone?

In my case I need input and output to be 5+6 for it to work as I’m intending, running on my iPhone connected via USB to the HX Stomp.

This way I can use cabs and post-FX on the Stomp, and run my pedalboard in front of a NAM capture. I’ve tested this with the vanilla NAM plugin in Logic Pro (standalone didn’t work on MacOS).

Also will my vanilla-trained captures from ToneZone3000 work in NAM XT? Grateful for any confirmation before I jump in!
 
Well, you're the one who couldn't hear the high frequency roll-off on tonex, so... I'll leave it at that :p
Wait, are we accepting that there is some sort of very audible loss of high end in Tonex captures? Should I be expecting an apology from all those gaslighting me for it even though I at least provided recorded files and graphs to back me up?

All that said, I still really enjoy the Tonex captures, I just need to tweak them a bit after capture
 
I’m looking at NAM XT for use with an HX Stomp. Could someone confirm if it’s possible to set whichever input and output on your audio interface in the standalone app on iPhone?

In my case I need input and output to be 5+6 for it to work as I’m intending, running on my iPhone connected via USB to the HX Stomp.

This way I can use cabs and post-FX on the Stomp, and run my pedalboard in front of a NAM capture. I’ve tested this with the vanilla NAM plugin in Logic Pro (standalone didn’t work on MacOS).

Also will my vanilla-trained captures from ToneZone3000 work in NAM XT? Grateful for any confirmation before I jump in!

Your captures will work - I can't answer the stomp question though
 
Wait, are we accepting that there is some sort of very audible loss of high end in Tonex captures? Should I be expecting an apology from all those gaslighting me for it even though I at least provided recorded files and graphs to back me up?
I've always been of that opinion... And my last capture of a drive pedal shows that incontrovertibly, to the point that I consider that capture unusable.

All that said, I still really enjoy the Tonex captures, I just need to tweak them a bit after capture
#MeToo
 
Let’s face it :
- super input actually reduces the aliasing to approach the best modelers oversampling levels
- but at a cost of TREMENDOUSLY more training time and recording
- Steve A doesn’t mind at all, and most of the community is thinking the same except DLC86, Francois and some scientists lost int the group
- aliasing at those levels is impossible to ear, even with A/B testing
- Francois is now banned of the nam group (and it was a great contributor, drama queen dude but funny) and he now tries to make his own NAM HD 96KHz (will need a 4k$ pc!!) in a new group on Facebook. That will probably ends up with nothing but some obscure graphics.
- nam is great but it’s an open source free plugin and it’s not as reliable as ToneX or Quad Cortex, and it’s NEVER be at the level of an axefx for the quality of amp. Steve is not building a team so it will also end up like a student project.
I wait for kemper NN or new big names bringing true innovation, but also A GOOD PRODUCT.
For now nothing beats Line 6 Helix and ToneX because they offer plugin + pedals and have a huge user base.
NAM is more like the 2000’s free metal plugins I played with. It’s cool for practicing at home.
 
Let’s face it :
- super input actually reduces the aliasing to approach the best modelers oversampling levels
- but at a cost of TREMENDOUSLY more training time and recording
- Steve A doesn’t mind at all, and most of the community is thinking the same except DLC86, Francois and some scientists lost int the group
- aliasing at those levels is impossible to ear, even with A/B testing
- Francois is now banned of the nam group (and it was a great contributor, drama queen dude but funny) and he now tries to make his own NAM HD 96KHz (will need a 4k$ pc!!) in a new group on Facebook. That will probably ends up with nothing but some obscure graphics.
- nam is great but it’s an open source free plugin and it’s not as reliable as ToneX or Quad Cortex, and it’s NEVER be at the level of an axefx for the quality of amp. Steve is not building a team so it will also end up like a student project.
I wait for kemper NN or new big names bringing true innovation, but also A GOOD PRODUCT.
For now nothing beats Line 6 Helix and ToneX because they offer plugin + pedals and have a huge user base.
NAM is more like the 2000’s free metal plugins I played with. It’s cool for practicing at home.
Considering that the literal point of open source software is that anyone can decide at any time to fork the repository and stand on the shoulders of the proverbial giants that came before them to build something in their own vision, I’m failing to understand the point of all this bitching and bellyaching.

If you disagree with the way the project is being run then fork the code, build your own team, put your money where your mouth is, and show some grit. This just reads like entitled BS.
 
Steve is not building a team so it will also end up like a student project.
I wait for kemper NN or new big names bringing true innovation, but also A GOOD PRODUCT.
For now nothing beats Line 6 Helix and ToneX because they offer plugin + pedals and have a huge user base.
NAM is more like the 2000’s free metal plugins I played with. It’s cool for practicing at home.
Something that often gets underestimated is just how different people's needs are with all this tech. Spark can be seen as the worst amp ever, and to others its a gamechanger (both are true for the right person).

For my capturing needs / anyone who's working ITB for personal or studio work, NAM hands down wins. The best quality captures "right now", No cost of entry / no barriers for literally anyone, quick reamp times, TZ3000 makes quick processing times, VST can be loaded across a whole session with multiple instances. Arguably one of the best open source / "free" technologies to ever come to guitar/pro audio. No other paid or unpaid solution is getting you better captures than NAM.

It wouldn't surprise me if Steve is consulting or working with some of the big names, so the tech you're waiting for is probably only a stones throw away from what NAM is doing now. If anything it'll basically be the same thing but just packaged within the L6 or [insert brand] ecosystem.

I love how some people are viewing Steve's movements as detrimental to NAM for some reason... like the guy who built it from the ground up and understands the tech better than anyone else can't possibly understand what it needs to move forward 😂. What would people have preferred?, Him scaling up a team and NAM becoming another bluguitar... taking a punt on a new company building out hardware and an entire ecosystem? Instant subscription models to suddenly fund this team of people working at NAM? All of his high level movements massively make sense to me, all of this is no surprise. I think if anything the naysayers will be surprised with what unfolds with NAM.
 
Let’s face it :
- super input actually reduces the aliasing to approach the best modelers oversampling levels
- but at a cost of TREMENDOUSLY more training time and recording
- Steve A doesn’t mind at all, and most of the community is thinking the same except DLC86, Francois and some scientists lost int the group
- aliasing at those levels is impossible to ear, even with A/B testing
- Francois is now banned of the nam group (and it was a great contributor, drama queen dude but funny) and he now tries to make his own NAM HD 96KHz (will need a 4k$ pc!!) in a new group on Facebook. That will probably ends up with nothing but some obscure graphics.
- nam is great but it’s an open source free plugin and it’s not as reliable as ToneX or Quad Cortex, and it’s NEVER be at the level of an axefx for the quality of amp. Steve is not building a team so it will also end up like a student project.
I wait for kemper NN or new big names bringing true innovation, but also A GOOD PRODUCT.
For now nothing beats Line 6 Helix and ToneX because they offer plugin + pedals and have a huge user base.
NAM is more like the 2000’s free metal plugins I played with. It’s cool for practicing at home.
Absolute garbage take, sorry to be so upfront. NAM is considered among the entire community the best in class, and will remain so for long, long time, because setting up the hardware is expensive and will hurt financially, except if some big player comes along and smahes processing power until no end in their next generation modeler. But even then, NAM will perhaps more likely will be embedded as an industry standard than anything else.

By the way, Francois was not a main contributor, but actually a very toxic and unknowing person spitting his misinformation around. He had this weird perception of anti-aliasing being the absolute worst enemy, which is total bullshit. He spammed multiple groups with his warfare against aliasing, and while at first I thought there was merit to it, but it was in the end cringe-inducing. Some of the others interested in reducing aliasing actually distanced himself it seems. He tried to ban me from his Neural Amp LABS group, to, guess what, I was reinvited.
But the worst was that he posted every.single.step in any group he could find. Even in the NAM Player group, which is totally unrelated, as the NAM Player seems to do it's own conversion of the NAM file, leading again to that weird artefacts being reintroduced.

I myself made a video in his superinput, and given what a weird dude he is, I am seriously considering removing that video from my channel, as I want no affiliation with this person, and to be fair, it is only slightly more efficient in reducing aliasing, but brings a whole can of worms (possible latency issues, longer training time, more customization, deviation from standards which are well established).

Francois was a hack, and it's good he got cast out of the community. By the way, he is constantly creating more and more profiles, like Lucy Stein, and on a forum he also posts as Audio Transformer, I heard...

... ... ... WAIT A MINUTE! ;-)
 
Last edited:
30 Rock Fellow Kids GIF by Peacock
 
Wait, are we accepting that there is some sort of very audible loss of high end in Tonex captures? Should I be expecting an apology from all those gaslighting me for it even though I at least provided recorded files and graphs to back me up?

All that said, I still really enjoy the Tonex captures, I just need to tweak them a bit after capture

Yeah, I think most high gain players immediately recognized it, but it was and is still great technology. I thoroughly enjoy playing through my ToneX pedal from time to time, especially the FX are great.

1740095807411.png
 
Yeah, I think most high gain players immediately recognized it, but it was and is still great technology. I thoroughly enjoy playing through my ToneX pedal from time to time, especially the FX are great.

And TBH that's actually not the worst thing for live playing anyway - sometimes I thought it sounded better as a result (at high volume). My tonex is still my backup swap-in for the dimehead if it goes down
 
I'm a simple man. I see a distortion/fuzz/overdrive pedal or a modeler, or an amp sim, or NAM plugin, or Tonex gizmo, and then enjoy the many awesome ways to make noise. This is the way. Noise. Turn it up and annoy your neighbours.
 
Back
Top