Alex Hofmann retweetledi

Stable Audio Open paper is finally out! 🎤
- High-quality stereo sound synthesis at 44.1kHz
- Can run on consumer-grade GPUs
- An autoencoder is implicitly released with the model
arxiv.org/pdf/2407.14358

English
Alex Hofmann
120 posts

@Alex_Hofmann_
Music researcher and performer. New project: "Études for live-electronics", mdw - Department of Music Acoustics (@iwk_mdw, @mdwwien)
























be careful which library you're using to load large numbers of audio files. librosa and pydub are very slow. the others are good. audioread is *very* fast.