Neural Amp Modeler alternative for the DWARF

I just discovered this open source project of a amp modeler capture.

I was wondering if this could ever being implemented in the MOD Dwarf as a plug in. I apologize in advance for my ignorance but would this be possible? This would be a complete game changer.


Just head on over here New neural lv2 plugin from Aida DSP based extensively on existing NeuralPi, reduced to the bone

coming soon…


amazing ,thanks for the link dude

This Neural Amp Modeler seems to be even quite good. There is a test of a guy available on YouTube:

And: Since ToneX the modelling is a hype topic. For Mod that is also a chance to place itself well in the hype.


the MOD should really focus on release a plug in based on that ASAP, that would put the device in a very good possition to compete with ToneX


Absolutely! ASAP! …can’t wait! :slight_smile:
I got the Tonex pedal, but the NAM (Neural Amp Modeler)
seems just as good.
Go for it!


This could be a great feature!
I tried NAM last week and is impressive!

1 Like

I’m not really the biggest fan of capture technology (for philosophical/workflow reasons) but it seems I’m in a small minority. Capture tech has arguably as much hype behind it as IRs did (and continue to have) back when they were getting popularized.

An open source tech that’s arguably on the level of proprietary solutions (Kemper/QC/ToneX) has to be one of the most exciting things happening in the guitar world right now. This is the kind of thing that still makes me check this forum from time to time, even though I returned my Dwarf more than a year ago. The main strength of Dwarf was always its potential, and while I didn’t feel like it delivered at the time, the plugin selection is now much stronger, and a thing like NAM can be the final difference maker for many potential customers.

I know MOD team has their hands full, but this might be the kind of thing that’s worth doubling down on.


I’m not that kind of nerd, so I need a plugin but is possible!

Today everything can profile or read a profile, here in Italy everyone talk about tonex pedal (can read only profiles)

1 Like

I’m genuinely curious about how many consumers of these devices (Kemper/QC/ToneX) buy them because they want to capture gear they have vs. those who buy them because they want to have a bunch of high quality simulators available in a convenient form factor. More specifically, how many of these consumers will basically never use the capture feature? Personally, I’ve always gravitated toward digital modellers because a) I didn’t want to spend several thousands on a bunch of gear b) I didn’t want to store and manage a bunch of gear I might only use very sporadically and c) I want the freedom to dabble in many different genres and styles.

I understand the sentiment, but I think MOD has to tread carefully here. The aforementioned companies have already spent years developing their technology and (importantly) user experiences. It’s not just that the capture technology could be on par with the current market offerings, but they’re also including comprehensive sets of high quality presets which take significant time and expense to curate. Headrush just released their new Prime model and initial reviews seem very positive, so the field is already getting more crowded with established players. If MOD throws their hat into this hype ring, the quality and user experience has to be competitive with existing products otherwise a few mediocre reviews might be enough for the general audience to write MOD off as sub-par despite the myriad capabilities that set the device apart from what the other products can do.

To be clear, I don’t mind that MOD doesn’t have the best simulators out there because I can do so many other things with the device that the others have no comparison for. I’m supportive of the idea of including a new plugin with this capability as “yet another awesome feature”, but I’m wary of MOD trying to compete with other products for whom this is a flagship part of their offering.


Yes, and MOD likely doesn’t have the resources to do the same. MOD, however, has the open source angle that makes them unique, and NAM is open source, which means that MOD doesn’t have to develop their own solution, they just need to incorporate NAM if possible (CPU limitation might be one of the first major barriers to overcome).

“The captures made by IK are not the greatest” was one of the main themes in most of the ToneX reviews. Default presets in modelers are a meme in general.

The point about UI and polish I do agree with, but this is more of an overarching MOD issue, rather than something that would be limited to a single plugin.

As opposed to the current situation, where Dwarf is universally considered a highly-polished, major contender in the modeler market, moving as many copies as the competition :grinning: ?

I’m of course being facetious for comedic effect… and I’ll continue in order to finish this sentence: in order to write something off as sub-par, you first have to actually know about it :grinning:

I’m not saying this from the position of authority or a market expert. I don’t even like amp capture technology that much, but it seems like people can’t get enough of it, and that could be Dwarf’s ticket to a wider appeal.

The bottom line is this: I’ve noticed a rare phenomenon of an open source software (NAM) actually getting quite a lot of traction, activity and hype behind it in a very short amount of time. People are sharing their amp captures, the Facebook group is growing fast, Nerds* are posting them “AB and null tests” on YouTube, suggesting that NAM can indeed “hang with the big boys and such”. I haven’t looked into any stats, but I’d bet dollars to doughnuts that NAM, a single open source plugin, is currently doing a lot better than MOD when it comes to social media engagement.

*I’m saying this with all due respect

I’m basically procrastinating writing this, I have no idea about anything. I don’t even like amp capture technology, at least not yet (despite trying both ToneX and NAM, both of which I liked but neither blew my mind), but people seem to love it for some reason, hence my uneducated suggestion.

1 Like

During the reboot, Gianfranco wrote a lengthy post about how MOD came to believe that being too closely associated with open source was a liability in establishing retail relationships and apparently led to many retailers being hesitant to work with MOD. Their fears - whether valid or not - are that having an open platform devalues the product because knock-offs are much easier to make. For example, while some of us may view the existence of Blokas / MODEP as a plus, retailers apparently do not. I could see similar difficulties coming if there were conversations on the internet about “well I can just run this on my Raspberry Pi or laptop”.

This sidesteps the point about needing to provide some set of relatively polished simulations. The tone snobs are always going to find some fault with the included captures, yet these devices are still selling. If the quality of the ToneX options isn’t good enough for the general consumer, another one of the options might be. You can be sure that Kemper, Line6, Headrush, et al will continue trying to raise the bar in quality. What would MOD ship with that wouldn’t cannibalize the existing paid options in their plugin Store?

MOD definitely has an identity crisis, and I wasn’t trying to suggest that they are doing well in that market segment, or any other segment for that matter.

It’s a good thought for sure. IMO marketing isn’t a strong suit for MOD and they could potentially benefit from any low-effort exposure and/or being included in conversations. One downside to this particular cohort is that in my experience, they tend to be notoriously cheap. Or, perhaps more generously, they tend to be more DIY and willing to roll their own. Maybe overgeneralizing, but many of these folks are the ones who come to forums and say “why should I pay $500 for something I can build in a weekend on a Raspberry Pi?” These might not be folks who are looking to spend money on a new product; rather their motivations are to achieve what others are doing without having to open their wallets if possible.

Neither do I, and I appreciate the exchange of ideas. My feeling is that MOD is quickly approaching another Hail Mary moment and this could indeed be the kind of initiative that propels them forward. My reservations come from the feeling that MOD is often running in 50 different directions and spread too thin to execute well on the big plays. Going in this direction feels like a “bet the house” kind of strategy - maybe that’s what needed?


Well-written reply, @unbracketed!

I would have agreed to this three months ago, but my perception is, that Mod Audio is doing really great currently. There are regular SW releases, communication about the expression pedal is forwarded to us, new pedals and plugins are published and the community seems to be thriving (although I’ll certainly miss @QuestionMarc’s comments). MODspeed, I say! :smiley:


I feel the same. The Mod team seems to me more focused on important things as before. There were new and very specific upgrades, lots of plugins went from Beta to Alpha etc. I felt the same as @unbracketed before but now I think Mod is much more fucused and on the right path.


Thanks a lot for the comment @eggsperde and @Kim . It really means a lot :slight_smile:

We’re indeed making a lot of effort to focus and improve the quality of deliveries.


I agree we did a lot of that in the past, but do you think that it has not changed? I’d be very happy to get your honest impression so that we can improve.

Given the traction this topic (AI assisted modelling) has gotten lately, I confess I have considered doing such a bet. But not yet :slight_smile:

What I can disclose is that we have something very exciting along those lines cooking in the oven :smiley:


I am not suggesting MOD to put all focus on the neural capture plug ins, just to do everthing they could to help the developers of those plug ins to get the most of it (helping testing it and giving advice how to optimize it). I am not a developer and probably i am talking rubbish lol. Anyway, I am veryexcited to see if those kind of plug ins go ahead. If that’s the case I believe the MOD DWARF would be all I need on my pedalboard. :slight_smile:


Reaction to this topic aside, the NAM group on Facebook just hit 3.5k users. They were celebrating 3k a day or two ago if memory serves. I’m not really an open source fanatic, more of a very casual open source enjoyer (I like the idea behind it) and to my casual eye, this kinda looks unprecedented, at least in the realm of guitar-related-audio-plugins.

I’m not a business person, I know nothing about the actual reality of selling a product like Dwarf. I’d just be making recommendations based on preferences, and uneducated beliefs, so I’m not saying anything like “you have to jump on it! this has to be the top priority!”. I just call it as I see it, through the lenses of a person who invested waaaaaaay too much time into the world of guitar gear and plugins to ever quit. :grinning:

However, it seems like the hype behind NAM is real, and for a good reason. I’m not a huge fan of gear captures, but people can’t get enough of them and there are enough sophisticated users out there, to produce some semi-objective (null tests and such) very competent comparisons of available profiling solutions.

It looks like among the most popular ones QC and Kemper are deemed “good”, ToneX and NAM are like half a tier higher (with NAM perhaps edging out ToneX slightly in accuracy). Also, folks “in the know” are already dismissing the profiling in the newest Headrush Prime as an inferior afterthought. Line 6 hasn’t entered the game yet. Fractal has their EQ match and pride.

Also, no shade on Aida DSP folks. I’m sure they worked hard on their stuff, and I have no way of testing the quality of it (even if I could, I’m hardly the most qualified person to do so), but assuming the capture tech clears the bar of being “good” then it lives or dies by the community willing to capture the gear they own. NAM got into my social media feed without me looking for them, whereas I have trouble finding anything of substance related to the Aida DSP amp profiling even when making a conscious effort. I’m sure this is unfair on multiple levels, but it’s also, you know, true.


Finishing products is hard, and I want to give credit, where credit is due. You’ve managed to keep the company afloat, and I can only guess how many fires you had to put out to get there. Enjoy your weekend!


@falkTX i see this has been added to mpb, does it build yet or is it still preliminary? I get a build error on filesystem.h, doesn’t seem to error out on any other builds…

it already works, but for duo and duox you need to use the -static variants.

just a few final details to handle, then it is up to aidadsp folks to decide on when to publish.
I am not even sure if I am supposed to share such screenshots yet anyway :sweat_smile: