Description

Uses neural networks to draw audio waveforms one audio sample at a time. Morph between different instruments in a grid.

Blog post:

https://magenta.tensorflow.org/nsynth-instrument

Demonstration video:

https://www.youtube.com/watch?v=rU2ieu5o5DQ

Hardware:

https://nsynthsuper.withgoogle.com/

Old Download Link:

http://download.magenta.tensorflow.org/demos/nsynth/NSynthProject.zip