DawDreamer

DawDreamer is an audio-processing Python framework supporting audio playback, VST instruments/effects, parameter automation, time-stretching, pitch-warping, and close integration with Faust.

DawDreamer was included as a Late-Breaking Demo at the 2021 ISMIR Conference. You can view the poster and read the paper.

One recent feature is the ability to transpile Faust code to JAX for machine learning tasks. I contributed the JAX backend to Faust and then demonstrated how to use it with DawDreamer via several Google Colabs [1, 2, 3].

🎛️Audio ML friends, I made a transpiler from Faust code to JAX! Faust is a powerful language for sound synthesis, so connecting it to JAX while supporting differentiability and learnable parameters is a huge research opportunity. 🧵 pic.twitter.com/qGQUeaQd7h

— David Braun (@DoItRealTime) October 20, 2022

I explained research opportunities involving Faust, DawDreamer, and AI in a talk at Faust Day 2023:

In December 2023, I did a talk and workshop on Faust-JAX at the Programmable Audio Workshop.

DawDreamer also supports time-stretching and beatmatching just like Ableton Live.

DawDreamer now supports BPM automation and comes with this demo script. Transition from song A at 128 BPM to song B at 132 BPM, back to A at 128. Many other new features too!https://t.co/HRLKKPfwWp

— David Braun (@DoItRealTime) May 21, 2022

DawDreamer has gained many features recently including pip install. A new notebook shows how to load Ableton warp marker files like this video. Faust integration enables custom polyphonic instruments. Hopefully very useful for ML researchers and artists. https://t.co/1ZJV2u49NI pic.twitter.com/F3IwynpdX6

— David Braun (@DoItRealTime) September 13, 2021