We have two clients (graphical and command-line) available for people to use to submit audio features to the AcousticBrainz database. These clients calculate a JSON file containing data about the audio and upload it to AcousticBrainz. We do not upload your audio.
All of these downloads are self-contained and include a static binary to perform the audio analysis. See our FAQ for more information on why we use static binaries.
To submit audio to AcousticBrainz, files must be tagged with MusicBrainz Identifiers. If they are not tagged, we recommend Picard.
You can build your own version of the submission tools. Download one of the following clients and follow the instructions in the README file:
Part of the source installation requires an extractor binary. You should use one of these static builds for Linux, Mac & Windows (see our FAQ why):
We provide regular dumps of the data in the AcousticBrainz database. The dumps are separated into low-level and high-level archives. See sample data for more information about the two data types.
Files in each archive are named according to the following structure:
type/mb/id/mbid-n.json
type
is one of lowlevel or highlevel,
mbid
is a uuid of a MusicBrainz Identifier, m
,
b
, i
and d
are the first, second,
third and fourth characters of the MusicBrainz
Identifier, and n
indicates the ordinal number of duplicate
data files present for the same MusicBrainz Identifier. There will always
be a file with number 0.
Released on January 30, 2015