acoustid-match
This is a simplified demonstration of the AcoustID algorithm, packaged as a Django app.
It computes fingerprints on a music collection, identifies similar tracks, and can optionally look up track artist and title via AcoustID.
Prerequisites
- A PostgreSQL database with the AcoustID Postgres Extension,
- The Chromaprint
fpcalcprogram. You can download it from acoustid.org. - A working Django site running on Postgres.
To look up fingerprints against the MusicBrainz database, an API key is needed in the ACOUSTID_API_KEY environment variable.
See here for instructions.
Usage
Add fingerprints to INSTALLED_APPS in the Django settings. Then run ./manage.py scan path/to/your/music/collection to add tracks. fpcalc is expected in the system PATH. Because fingerprinting might take a while for a large music collection, scan uses all available CPUs in parallel.
When completed, head to the Django admin site. Your music is visible under fingerprints, duplicates are already marked. Use the Identify with MusicBrainz action in the track list to look up artists and titles of unknown tracks. If successful, tracks are marked as Identified and the JSON data is shown in the fingerprint details.
Credits
Kudos to Lukáš Lalinský, the author of the excellent AcoustID software.