Let's Write a Reverb (2021)

For reverb I don't see much practical use, mainly because you can capture a pretty-much-perfect recreation of a real space with an impulse response. No need for thousands or millions of rounds of training a network. For unrealistic reverbs, you have the problem that to get training data you'd have to invent several unrealistic reverb effects to apply to sounds. And once you've made those effects, there's not really any reason to neural-netify them.

For NeuralDSP it's a bit different because they use NN's to simulate a guitar amp circuit which is a nonlinear system and so there's no simple way to "capture" the effect the way that you can for reverb sims or speaker sims. And while you can make a very accurate model using something like SPICE, that won't run in realtime. With traditional amp modeling you basically take the SPICE version and try to optimize and cheat as much as you can so it can run in realtime, at the cost of accuracy.

So that's what NeuralDSP's goal is, a system that approximates the amplifier but can also be computed in real-time, except done using a trained NN instead of a human-optimized variant of the SPICE circuit.

They have a couple whitepapers on their website, though none of them go deep enough to really give away their secret sauce. But basically according to them, making a NN model of an amplifier at a fixed setting is fairly simple. Where they had to get novel with it is adjustable settings/parameters. E.g. turning the drive up, or turning the treble down. Just capturing a few hundred or thousand models based on adjusting parameters and cross-fading between them doesn't sound realistic. So they had to come up with a larger model architecture that can "learn" those parameter changes.

https://arxiv.org/pdf/2403.08559

Fantastic article from start to finish. Great explanation, great audio samples, and from a first skim the C++ example looks very readable.

> I haven't found any good resources on this particular diffuser design. I found a couple of forum posts and a paragraph from a book

When I read this my first thought was "this is a link to a comment by mystran on KVR, isn't it?" And yep. If you're looking for some obscure DSP knowledge, all Google searches eventually lead to a comment by mystran.

Very nice tutorial, and extremally nicely done!

It's worth knowing that some of those effects can also be achived (of course much much more simplified) in all modern browsers using the Web Audio API. I created mobbler[0] using that, and I also wrote a small tutorial on how some of the effects can be achieved using simple modules (it might seem too complex at first glance, but you can just look at the pictures)[1].

[0]: https://github.com/Megaemce/mobbler

[1]: https://github.com/Megaemce/mobbler/wiki/Tutorials

It is 100% possible and there are a slew of tricks you can use to get big performance boosts with negligible cost to accuracy.

It’s not that hard, you just collect a lot of data. Much easier with a robot turning the knobs. Predict the next sample based on input and knob settings.