Is there a reason why there are lots of different knobs for the plug in, like the EQs and pregain, etc.? I was assuming that the neural net itself is only a snapshot, with the knob settings not given as input. I guess it’s just for convenience?
Without such controls, those profiles would be too limited as “one trick ponies” and you have to try to tweak it out after its master volume.
I do wonder how presence (and its opposite “depth”) will work on those. Is it a certain EQ range that gets cut/boost? Is it the same Freq range as the depth and presence on the Veja Mutant?
we’re preparing doc for this with everything explained. Presence is an high-shelf, depth is a parametric boosting low end.
Will there be a possibility to adjust this for bass players?
this will also work for bass (and pedals and most other devices). I’m training a few models for that but they will need a little more time.
excitement++;
Hi everyone
A quick bump here.
AIDA-X is officially released: MOD presents the AIDA-X and dives into Neural Modelling
Not exactly “betting the house” @unbracketed, but we are definitely throwing some serious chips on this table
The timing is remarkably good and this topic itself is scoped to a point that we can all make a tangible effort, including the community members. I would like to use this opportunity not only to tap into external communities that cater to our product, but also to reorganize how our community itself organizes and moves forward.
More on this in the next days.
I hope you guys are excited as we are!
I don’t want to pretend I know better than you on your products, but I’d like to be optimistic…
NAM apparently works well on R-Pi4 (quad 1.4GHz), which has a similar processing power as the Dwarf (quad 1.3GHz), but half of the processing power of a DuoX… but I’m not a coding person, so my skills are limited to support you guys, and say thank you for your efforts! MOD is amazing, keep going!
Well there has been some improvements, we are now able to load a few NAM models on a Duo X unit but not on a Dwarf yet
| What’s happening there? Are you guys aiming at making NAM proliles work on MOD devices?
Hope I don’t pass the wrong idea, I am not that much involved but have been talking to them because it is another nice open-source based project.
For curiosity I check the NAM side from time to time, to see if there were any changes regarding performance and general behaviour. They seem interested on trying an LSTM approach similar to AIDA-X, which is curious.
There is not much to see at all yet. On mod-ui side I already made it so that “nammodel” is an acceptable file type, just to be able to list NAM files (requires manual upload via SSH).
But due to WaveNet models being so heavy on resources in general, I don’t expect us to be able to run them as-is any time soon. Would be happy to be proven wrong of course
If you manage to have NAM run smoothly (like AIDA-X) on your units, you guys will have only one problem: production, to match the demand, because every guitarists nowadays are asking for the same thing: a merged fx/modeler pedal with quality plugins, that also can load amp tone captures.
That is what you guys are offering with the Dwarf+AIDA-X, but as you know, for any open source project, the size of the community is key. As you very well know, NAM already has a very strong community… If you manage to do that, it will be a dead ringer for a famous (overhyped) competitor’s pedal… and for guitarists, MOD will become a one stop shop.
It might be worth thinking at re-issuing a powerful ground unit like the DuoX (or even more powerful) at Dwraf’ price point, with an official NAM running on it (yes, I know… the ‘great food for a cheap price’ concept is not easy to accomplish, but…). I don’t know… get the NAM’s guy to work actively in helping you, and give him some shares of the company if you cant pay him right now… NAM is really an added value for the MOD, imo. Anyways… just my 2 cents suggestion.
Good luck, and thanks again!
the platform is open to such plugins and more, but I am afraid it is not just a matter of putting hours into it.
the wavenet based models, as used by NAM, are just heavy by design. there is progress to be made still in terms of optimizations but (from what I can see) there is a clear ceiling that can’t be overcome by pure optimizations alone.
it sometimes is just a question of luck too.
even with all the technical issues around NAM, the plugin got a big surge of popularity that the authors were not expecting.
A question of youtubers more so than luck… it’s sad to say, but in order to have any people buy the same toy, influencers have to promote it (aka get it for free…)
I think the success of NAM is the fact that its interface doesn’t need any proprietary device to work on computers. Whichever company will integrate NAM to their fx/mod pedal will win the game, imo. If optimization is no option, then higher processing power is the way to go (I know… it has a cost).
Alternatively, another way to ‘conquer’ ground (in order to gain time and create a larger online community while optimizing it for a future Mod pedal), would be to offer a computer-only MOD gui for a reasonable price to use on pc (without the need of Dwarf/DuoX, just a generic midi interface), and there offer NAM plugin (that uses the computer’s CPU). The gui (even without AidaX or NAM) is really unique. People pay for Amplitube, no reason why they wouldn’t for MOD’s gui.
And that would also help diversify products, and thus increase income (before some Linux dev decides to create a MODEP for regular computers…). Again, just my 2 cents.
mod-ui and mod-host are fully opensource and already can run on your desktop machine (although pretty much exclusively on Linux I think).
Also there’s the live-usb that @Rom just linked
Yes, exactly. Thanks for putting that one my radar, sorry I’m a real MOD-newbie
So in order to run that, I first have to install ArchLinux, if I get it right. First time I hear about that thing…
I had a look on their platform, where they state “if you are a beginner and want to use Arch, you must be willing to invest time into learning a new system”. And that is the reason why a community doesn’t grow: they speak a different language (I’m not talking about R or Python, I’m talking about the English terminology in the installation guide…)
Is there a tutorial understandable by mortal Humans I could find somewhere? A youtube tutorial maybe like ‘Running MOD as an USB image - form zero to hero’? I’m willing to invest a bit of time understanding that ArchLinux OS (and that is more than 99.9% of what most will agree to do before their 3min34sec attention span vanishes), but keeping in mind that my goal is to make new sounds with my guitar, not to get a PhD in computer science.
Thoughts?
Nope, you just need a USB key, you flah the latest .iso with a software like Etcher, then you boot your PC on the key (absolutely no install needed)
(All this seams written in the github page)
Yes, you’re correc! Sorry for having asked but the only useful 1-liner on the tutorial page was sunk in the middle of lines of codes and instructions for developers… very confusing.
I spent a lot of time digging on the websites and forums to find that answer, it’d help develop the MOD community to put a link to that on the front page of the MOD website so that many can use it without the need of a Duo/Dwarf. That said, I also understand why they would not make it easy to find that solution…
Thank you for your help!
you know aida-x has a desktop plugin version right?
I did some of the work there. this plugin works for as many plugin formats as possible - LV2, VST2, VST3, CLAP and AU, also a standalone. plus for convenience it can load a cabsim IR within the plugin too.
Getting all the MOD stuff to work directly on PCs would for sure increase its popularity, but I do not see how it would increase hardware sales (I think it would even reduce it, since people wouldnt need a Dwarf as much), and in the end this would hurt the company more than help.
PS: The live-usb is simply a research project I did on my own, not even during work hours. It is one of many side things done to experiment with, then show off to the rest of the team to see if it could be useful in some way. Typically in most companies such projects wouldnt be open until there was an official release, but I dislike keeping things hidden even if not finalized.