MicroMusic is a free AI-powered patch generator for Vital that recreates synth patches from audio.
You input an audio sample, and MicroMusic outputs a Vital preset file. Behind the scenes, MicroMusic uses machine learning to find the optimal parameters to create the closest matching preset it can.
Here’s an intro to how MicroMusic works:
Availability:
MicroMusic is available now as a free download. The official release of MicroMusic is planned for June 1, 2024.
“In general you want isolated and simple sounds.”
It’s a sound generator… throw anything you want at it, think outside the box already.
Unfortunately windows only…
This is basically ML-fueled resynthesis.
The examples where it missed the mark still produced nice patches, which makes me think that adding some randomness/inaccuracy could be useful.
I wonder whether as a sound designer I’d be mad that AI was copying my patches or that I wouldn’t care because sampling has been a thing for decades.
Resynthesis has often been aimed at additive synthesis (which lends itself to a workflow of partial extraction and regeneration) or at multisampling. But applying it to other synthesis models (FM/subtractive/granular/wavetable/physical modeling/modular/etc.) seems like it could be very useful. It probably lends itself to statistical and machine learning approaches since you’re trying to generalize a function/model that maps audio into patch parameters, minimizing the audible difference between the generated audio and the input sample.