A fast, accessible audio player for Windows, macOS and Linux with support for tempo/pitch shifting, effects, and screen reader accessibility.
- Tempo, pitch, and rate adjustment (Speedy for speech, Signalsmith Stretch for music)
- Audio effects (reverb, echo, EQ, compressor, stereo width, center/vocal cancel, convolution, 3D audio)
- Recording to WAV, MP3, OGG and FLAC
- MIDI with SoundFonts, and tracker modules
- Internet radio streaming with favorites
- YouTube audio playback
- Screen reader support via Universal Speech
- Hotkey support
- Chapter navigation for audiobooks
- Visual Studio 2022 (or its Build Tools) with the "Desktop development with C++" workload
- Download from Visual Studio Downloads
- CMake 3.21 or newer
- Git (CMake fetches some dependencies with it)
Run download-deps.bat once. It downloads SQLite, Speedy, Sonic, KissFFT,
Signalsmith Stretch and FDK AAC into deps/ and src/, and FastPlay's FFmpeg
into ffmpeg/: the audio-only build CI makes with ci/ffmpeg/build.sh, taken
from CI's latest run with the GitHub CLI (gh auth login first).
wxWidgets (the user interface toolkit), UniversalSpeech (screen reader speech), miniaudio, SpessaSynth, libopenmpt, LAME, libogg and libvorbis are fetched and built by CMake itself on the first build.
Open a command prompt and run:
build_new.batThis configures CMake in build/, builds FastPlay.exe, and packages
FastPlay.zip (and FastPlayInstaller.exe when Inno Setup is installed). The
first build takes a while because it compiles wxWidgets; later builds reuse it.
Disable screen reader support:
build_new.bat no-speechInstall a compiler, CMake, Git and the development packages for GTK 3, libcurl, OpenSSL, zlib, Speech Dispatcher and ALSA. On Fedora:
sudo dnf install gcc-c++ cmake ninja-build git gtk3-devel libcurl-devel openssl-devel \
zlib-ng-compat-devel speech-dispatcher-devel alsa-lib-develOn Debian and Ubuntu:
sudo apt install g++ cmake ninja-build git libgtk-3-dev libcurl4-openssl-dev libssl-dev \
zlib1g-dev libspeechd-dev libasound2-devThen run ./download-deps.sh once (on Linux it builds FastPlay's FFmpeg itself,
which takes a few minutes), and build:
cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release
cmake --build build
build/FastPlayOn Linux FastPlay plays through PulseAudio, or PipeWire's PulseAudio service;
speaks through Speech Dispatcher, as Orca does; shows what is playing and takes
the media keys through MPRIS; and plays MIDI through output ports of the ALSA
sequencer (without ALSA's headers, through the built-in synthesizer only). Global
hotkeys need an X11 session: Wayland does not let a program take keys from the
others. Settings and the library are in ~/.local/share/FastPlay.
After building, run FastPlay.exe. The screen reader client DLLs are loaded from the lib/ subfolder.
FastPlay/
├── src/ # Player, effects, database and other core code
│ ├── audio/ # The audio engine: decoders, tempo, output, recording
│ ├── ui/ # The user interface (wxWidgets)
│ ├── platform/ # Operating system specific code
│ ├── reverb/ # The reverb engines
│ └── spatial/ # The HRTF behind 3D Audio (SADIE II data, pffft)
├── include/ # Header files
├── lib/ # Screen reader client DLLs
├── ffmpeg/ # FastPlay's FFmpeg build (ci/ffmpeg)
├── deps/ # Third-party dependencies (Speedy, Sonic, Signalsmith Stretch, etc.)
├── res/ # Windows resources
├── CMakeLists.txt # Build definition
├── build_new.bat # Build script
└── FastPlay.exe # Output executable
This project uses the following third-party libraries:
- FFmpeg (LGPL 2.1 or later), built from source by ci/ffmpeg/build.sh
- Codec 2 (LGPL 2.1), its speech codec built into FFmpeg by ci/ffmpeg/build.sh
- miniaudio (public domain / MIT No Attribution)
- SpessaSynth, C port by kode54 (Apache 2.0)
- libopenmpt (BSD)
- LAME (LGPL)
- libogg and libvorbis (BSD)
- Fraunhofer FDK AAC (FDK AAC licence)
- Speedy (Apache 2.0), Sonic (Apache 2.0), Signalsmith Stretch (MIT)
- SQLite (public domain)
- wxWidgets (wxWindows Library Licence)
- Universal Speech (MIT)
- SADIE II HRTF database, University of York (Apache 2.0)
- pffft (FFTPACK licence)