Spotify’s Pedalboard Brings DAW-Grade Audio Effects to Python

Spotify has open sourced Pedalboard, a Python framework that brings studio-quality audio effects to code. Built on top of JUCE, the industry-standard C++ framework for audio applications, Pedalboard lets developers apply professional effects without launching a digital audio workstation (DAW).

Music and podcast production today happens overwhelmingly inside DAWs such as Ableton Live, Logic Pro, or Pro Tools, as well as more accessible tools like Soundtrap and Anchor. These applications are optimized for high performance and audio quality, giving producers fine-grained control over their sound. But DAWs are built for musicians, not programmers. Engineers and researchers at Spotify’s Audio Intelligence Lab needed the power and speed of a DAW within Python code for their audio research, and found no existing package met all their requirements. Pedalboard is their solution.

Speed and Compatibility

Like a professional DAW, Pedalboard ships with a range of built-in effects and supports third-party VST3 and Audio Unit plugins. It also prioritizes performance: in basic tests on common developer hardware, Spotify reports Pedalboard runs up to 300 times faster than widely used Python audio effects packages.

The package works across macOS, Windows, and Linux out of the box. Spotify says it has processed millions of hours of audio with Pedalboard internally over nearly a year of use before the release.

Effects and Plugin Support

Pedalboard’s name reflects its guitar-pedal heritage: it includes common stylistic tools for altering sound, as well as foundational utilities for controlling volume. Built-in effects include:

  • Noise gate, compressor, and limiter for dynamic control
  • Distortion, phaser, filter, and reverb for stylistic shaping
  • A convolution operator for high-quality speaker and microphone simulation

When the built-in effects aren’t enough, developers can load any VST3 or Audio Unit effect plugin to expand the sonic palette. Effects can be grouped into a single pedalboard object, which not only saves the setup for reuse but also speeds up processing by chaining operations efficiently.

Use Cases at Spotify

Spotify has identified several practical applications for Pedalboard in its own workflows:

Machine learning. Pedalboard accelerates data augmentation dramatically while producing more realistic results. A small dataset can be expanded with reverb, compression, distortion, and other effects to grow training data and improve model performance. The library has been extensively tested in high-performance and high-reliability ML pipelines at Spotify, particularly with TensorFlow.

Content creation. Scripting effects application in Python makes it possible to automate parts of the audio production process. Applying a VST3 or Audio Unit plugin no longer requires manually opening a DAW, importing audio, and exporting the result; a few lines of code handle the entire operation in a single command or as part of a larger pipeline.

Creative experimentation. Artists and producers comfortable with Python can use Pedalboard to craft novel effects that would be extremely tedious to build in a DAW. For programmers new to audio, the library also serves as an approachable bridge between code and music.

Getting Started

Pedalboard is available now on GitHub, where Spotify welcomes contributions. Installation is a single command:

pip install pedalboard

The open source release follows a tradition of contributions from Spotify’s research labs, and the company hopes the library unlocks new possibilities for researchers, engineers, and musicians.

VST is a registered trademark of Steinberg Media Technologies GmbH.