DBraun/DawDreamer

Digital Audio Workstation with Python; VST instruments/effects, parameter automation, FAUST, JAX, Warp Markers, and JUCE processors

What it solves

DawDreamer provides a Python-based framework for audio processing, allowing users to programmatically control a Digital Audio Workstation (DAW) environment. It bridges the gap between professional audio software (like VST plugins) and Python, enabling the automation of audio rendering, composition, and machine learning experiments with sound.

How it works

Built on the JUCE framework and using nanobind for Python integration, DawDreamer allows users to create graphs of audio processors. Users can load VST instruments and effects, write Faust code for custom signal processing, and manage MIDI playback and parameter automation. It can render audio to files or play it back in real-time, with support for time-stretching and pitch-warping.

Who it’s for

It is designed for audio engineers, researchers in music technology, and developers who want to integrate professional audio plugins and signal processing into Python workflows, particularly for those conducting machine learning experiments with audio.

Highlights

  • VST Support: Load and control VST instruments and effects with UI editing and state management.
  • Faust Integration: Support for Faust effects and polyphonic instruments, including the ability to transpile Faust code to JAX/Flax.
  • Advanced Automation: Parameter automation at audio-rate and pulses-per-quarter-note (PPQN).
  • Flexible Rendering: Ability to render and save multiple processors simultaneously.
  • Warping: Support for time-stretching and pitch-warping, including Ableton Live warp markers.
  • Cross-Platform: Full support for macOS, Windows, and Linux.

Related

  • Project
    spotify/pedalboard

    A Python library for reading, writing, and applying studio-quality audio effects to audio files or live streams, supporting VST3 and Audio Unit plugins.

  • Project
    Eventual-Inc/Daft

    A high-performance, Rust-powered data engine designed to process multimodal AI workloads, allowing users to handle images, audio, and video alongside structured data at scale.

  • Project
    sdatkinson/NeuralAmpModelerPlugin

    A VST3/AudioUnit plug-in that integrates the Neural Amp Modeler engine into DAWs for AI-driven guitar amplifier modeling.

  • Project
    Noizefield/audio-plugin-coder

    An AI-powered framework that guides LLM agents through the full lifecycle of creating audio plugins, from initial concept to shipped VST3/AU/CLAP products.

  • Project
    meizhong986/WhisperJAV

    WhisperJAV is a local Windows/macOS/Linux tool that generates Japanese subtitles for adult‑video files. It builds a custom pipeline (scene detection → optional denoising → VAD → Whisper‑family ASR → Japanese‑specific post‑processing) to handle noisy, long‑form audio. Users can run a simple GUI or CLI, choose from several processing modes (balanced, fast, fidelity, Qwen, anime‑whisper, etc.), adjust VAD sensitivity, or run a two‑pass ensemble that merges results from two different pipelines. The software runs entirely on the user’s machine, supports translation via local or cloud LLMs, and provides installers and Colab/Kaggle notebooks for quick start.