Fugatto: Nvidia's new audio/music foundation model

In case you haven’t seen this elsewhere… Nvidia has unveiled an impressive new model for interacting with sound. Pretty mind-blowing. I think the implications will be profound…

6 Likes

The voice and singing generation is good - but the sound generation felt like it had a lot of artifacts in it to me - not the most impressive demos - cool technology though!

2 Likes

I’m hoping this will help fulfil my dream of a full multi-sampled dog bark sfz. Then we can finally put to rest literally the only thing wrong with General MIDI: the single dog bark sample.

2 Likes

“plus ça change…”

2 Likes

CSound for noobs (me). this is cool.