Retrieving MusicBrainz fingerprint with python

Question

I struggle to find a way to retrieve the audio fingerprint data from given songs. I tagged those songs with MusicBrainz before and now I want to read it directly from the audiofile. The only way I found with musicbrainzngs was to search (musicbrainzngs.search_artists()) for the song in the MusicBrainz library to find the fingerprint data which is rather inefficient since I have the data in my files already.

Is there a library where I can read it from any given audiofile?

Answer 1

The AcoustID fingerprint is stored in audiofiles, yes, if you tagged your files using MusicBrainz Picard. What tag exactly depends on the audio file:

• MP3 / ID3V2: TXXX:Acoustid Fingerprint
• Vorbis (FLAC, ogg) and APE: ACOUSTID_FINGERPRINT
• iTunes MP4: ----:com.apple.iTunes:Acoustid Fingerprint

See the Musicbrainz tag mapping overview.

You can read audiofile tags with the mutagen project, which is what Picard uses too:

import mutagen

tagnames = {
    'audio/mp3': 'TXXX:Acoustid Fingerprint',
    'audio/vorbis': 'ACOUSTID_FINGERPRINT',
    'audio/mp4': '----:com.apple.iTunes:Acoustid Fingerprint'
}

mfile = mutagen.File(musicfile_filename)
if mfile and mfile.mime[0] in tagnames:
    fingerprint = mfile.get(tagnames[mfile.mime[0]])
    if fingerprint is not None:
        try:
            # ID3v2 tag
            fingerprint = fingerprint.text[0]
        except AttributeError:
            # Vorbis
            fingerprint = fingerprint[0]