ReplayGain Plugin#

This plugin adds support for ReplayGain, a technique for normalizing audio playback levels.


This plugin can use one of many backends to compute the ReplayGain values: GStreamer, mp3gain (and its cousin, aacgain), Python Audio Tools or ffmpeg. ffmpeg and mp3gain can be easier to install. mp3gain supports less audio formats than the other backend.

Once installed, this plugin analyzes all files during the import process. This can be a slow process; to instead analyze after the fact, disable automatic analysis and use the beet replaygain command (see below).

To speed up analysis with some of the available backends, this plugin processes tracks or albums (when using the -a option) in parallel. By default, a single thread is used per logical core of your CPU.


To use GStreamer for ReplayGain analysis, you will of course need to install GStreamer and plugins for compatibility with your audio files. You will need at least GStreamer 1.0 and PyGObject 3.x (a.k.a. python-gi).

Then, enable the replaygain plugin (see Using Plugins) and specify the GStreamer backend by adding this to your configuration file:

    backend: gstreamer

The GStreamer backend does not support parallel analysis.

mp3gain and aacgain#

In order to use this backend, you will need to install the mp3gain command-line tool or the aacgain fork thereof. Here are some hints:

  • On Mac OS X, you can use Homebrew. Type brew install aacgain.

  • On Linux, mp3gain is probably in your repositories. On Debian or Ubuntu, for example, you can run apt-get install mp3gain.

  • On Windows, download and install the original mp3gain.

Then, enable the plugin (see Using Plugins) and specify the “command” backend in your configuration file:

    backend: command

If beets doesn’t automatically find the mp3gain or aacgain executable, you can configure the path explicitly like so:

    command: /Applications/

Python Audio Tools#

This backend uses the Python Audio Tools package to compute ReplayGain for a range of different file formats. The package is not available via PyPI; it must be installed manually (only versions preceding 3.x are compatible).

On OS X, most of the dependencies can be installed with Homebrew:

brew install mpg123 mp3gain vorbisgain faad2 libvorbis

The Python Audio Tools backend does not support parallel analysis.


This backend uses ffmpeg to calculate EBU R128 gain values. To use it, install the ffmpeg command-line tool and select the ffmpeg backend in your config file.


To configure the plugin, make a replaygain: section in your configuration file. The available options are:

  • auto: Enable ReplayGain analysis during import. Default: yes.

  • threads: The number of parallel threads to run the analysis in. Overridden by --threads at the command line. Default: # of logical CPU cores

  • parallel_on_import: Whether to enable parallel analysis during import. As of now this ReplayGain data is not written to files properly, so this option is disabled by default. If you wish to enable it, remember to run beet write after importing to actually write to the imported files. Default: no

  • backend: The analysis backend; either gstreamer, command, audiotools or ffmpeg. Default: command.

  • overwrite: On import, re-analyze files that already have ReplayGain tags. Note that, for historical reasons, the name of this option is somewhat unfortunate: It does not decide whether tags are written to the files (which is controlled by the import.write option). Default: no.

  • targetlevel: A number of decibels for the target loudness level for files using REPLAYGAIN_ tags. Default: 89.

  • r128_targetlevel: The target loudness level in decibels (i.e. <loudness in LUFS> + 107) for files using R128_ tags. Default: 84 (Use 83 for ATSC A/85, 84 for EBU R128 or 89 for ReplayGain 2.0.)

  • r128: A space separated list of formats that will use R128_ tags with integer values instead of the common REPLAYGAIN_ tags with floating point values. Requires the “ffmpeg” backend. Default: Opus.

  • per_disc: Calculate album ReplayGain on disc level instead of album level. Default: no

These options only work with the “command” backend:

  • command: The path to the mp3gain or aacgain executable (if beets cannot find it by itself). For example: /Applications/ Default: Search in your $PATH.

  • noclip: Reduce the amount of ReplayGain adjustment to whatever amount would keep clipping from occurring. Default: yes.

This option only works with the “ffmpeg” backend:

  • peak: Either true (the default) or sample. true is more accurate but slower.

Manual Analysis#

By default, the plugin will analyze all items an albums as they are implemented. However, you can also manually analyze files that are already in your library. Use the beet replaygain command:

$ beet replaygain [-Waf] [QUERY]

The -a flag analyzes whole albums instead of individual tracks. Provide a query (see Queries) to indicate which items or albums to analyze. Files that already have ReplayGain values are skipped unless -f is supplied. Use -w (write tags) or -W (don’t write tags) to control whether ReplayGain tags are written into the music files, or stored in the beets database only (the default is to use the importer’s configuration).

To execute with a different number of threads, call beet replaygain --threads N:

$ beet replaygain --threads N [-Waf] [QUERY]

with N any integer. To disable parallelism, use --threads 0.

ReplayGain analysis is not fast, so you may want to disable it during import. Use the auto config option to control this:

    auto: no