Corrupted AcoustIDs?


I’m merging Dolly Partons recordings. I noticed same fingerprints that does not have the normal length (e.g.

I normally disable then.

How it comes and is it alright to disable this?

I’ve had acoustid-fingerprinter & fpcalc generate broken AcousticIDs (often given weird complaints about perfectly valid FLAC files). There is surely a bug somewhere, haven’t spent the time to try chasing it down. Picard, OTOH, works…

[I’m running Debian testing. Also have deb-multimedia packages for some things, that might have to do with it.]

This AcoustID seems legit; it is from a file with the right length (3:21), and a comparison with another fingerprint of this recording shows they are quite similar.

Fingerprints normally consider the first two minutes of audio from a file. In this case, however, only a shorter portion (20 seconds perhaps) has been fingerprinted; why that is, I have no idea.

At the beginnings of AcoustID, they were shorter, like this.

They were shorter, yes, but not so short (1 minute instead of 2). Half of a “new” fingerprint.