Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?


Failed to load latest commit information.
Latest commit message
Commit time
July 20, 2015 00:06
August 28, 2015 16:53
January 16, 2023 23:20
September 14, 2015 20:18
November 15, 2015 01:42
August 8, 2014 12:35
March 16, 2016 16:37


This library provides decoding and encoding of audio on a playlist. It is intended to be used as a backend for music player applications. That said, it is also generic enough to be used as a backend for any streaming audio processing utility.


  • Uses ffmpeg for robust decoding and encoding. See supported file formats and codecs.
  • Add and remove entries on a playlist for gapless playback.
  • Supports idempotent pause, play, and seek.
  • Per-playlist-item gain adjustment so you can implement loudness compensation without audio glitches.
  • Read and write metadata tags.
  • Choose between smooth mode and exact mode during playback.
    • smooth mode - open the audio device once and resample everything to fit that sample rate and format.
    • exact mode - open and close the audio device as necessary in effort to open the audio device with parameters matching the incoming audio data.
  • Extensible sink-based interface. A sink provides resampling and keeps its buffer full. Types of sinks:
    • raw sink - Provides reference-counted raw audio buffers you can do whatever you like with. For example a real-time audio visualization. All other sink types are built on top of this one.
    • player sink - Sends frames to a sound device.
    • encoder sink - Provides encoded audio buffers. For example, you could use this to create an HTTP audio stream.
    • loudness scanner sink - Uses the EBU R 128 standard to detect loudness. The values it produces are compatible with the ReplayGain specification.
    • fingerprint sink - Uses chromaprint to generate unique song IDs that can be used with the acoustid service.
    • waveform sink - Generates a visual representation of a song.
  • Example programs included:
    • playlist - Play a series of songs with gapless playback.
    • metadata - Read or update song metadata.
    • replaygain - Report the suggested replaygain for a set of files.
    • transcode - Transcode one or more files into one output file.
    • fingerprint - Generate acoustid fingerprints for one or more files.
    • metadata_checksum - Read or update song metadata. This program scans the audio of the file before the metadata change, changes the metadata in a temporary file, scans the audio of the temporary file to make sure it matches the original, and then atomically renames the temporary file over the original file.

Building From Source


  • cmake >= 2.8.5
  • ffmpeg >= 3.0
    • suggested flags: --enable-shared --disable-static --enable-libmp3lame --enable-libvorbis --enable-gpl
  • libebur128
    • make sure it is compiled with the speex dependency so that true peak functions are available.
  • libsoundio
  • libchromaprint-dev
mkdir build
cd build
cmake ..
sudo make install


API Reference

Join #libgroove on and ask questions.

To build the documentation:

make doc

Projects Using libgroove

Feel free to make a pull request adding yours to this list.

  • Groove Basin is a music player with lazy multi-core replaygain scanning, a web interface inspired by Amarok 1.4, http streaming, upload, download and a dynamic playlist mode.
  • waveform generates PNG waveform visualizations.
  • node-groove provides Node.js bindings to libgroove.
  • playa OS X Audio Player that thinks in albums.
  • groove-rs provides rust bindings to libgroove.
  • ruby-groove provides Ruby FFI bindings to libgroove.
  • TrenchBowl is a simple Qt GUI on top of libgroove.


streaming audio processing library







No packages published