pedalboard

Why we included this project

Audio teams working in Python often have to leave the ecosystem to get studio-quality processing, and this library removes that step. Built by Spotify's Audio Intelligence Lab for data augmentation and machine-learning work, it reads, writes, and renders audio while shipping a solid set of built-in effects: reverb, compression, EQ, time-stretching, and more. Its standout capability is hosting VST3 and Audio Unit plugins, so you can load third-party commercial instruments and effects and run them inside a script rather than a DAW. That makes it a practical tool for ML pipelines that need realistic perturbations of training audio, and for batch jobs that apply the same treatment across many files. Teams doing speech, music, or general sound research will find it sits comfortably between quick experiments and production audio work.