I think the easiest solution for this wouldn’t involve Picard, but some scripting for running fpcalc and getting and embeding just the acoustid and the fingerprint from the script. This should be fairly easy using pyacoustid and mutagen, if you are a little bit familar with python. I think even the AI dev tools should be able to get you a good script to start with.
2 Likes
@dashv And one unrelated question, can i contact you in the future if i ever get to developing a more advanced system for fingerprint or some other musical analysis? Having access to such a large library would help a lot, if I could ask you to run some tools on it.
3 Likes
Absolutely, I can run batch in the background 24 hours a day 7 days a week (I7 4cpu 32GB memory), storage is no issue. I am starting to digitize my vinyl, for some albums I have both CD and Vinyl and wanted to look at the difference in the mastering of each, it would be nice to know how to do that other than visually in Audacity. (All those are later discussions).
As for the ID I already have the fingerprint, the only thing I am missing is the associated ID due to the way I process & tag. I played around earlier and realized I will need write a simple REST script to grab it. Been 10 years since I wrote one for a storage system so it will be fun getting rid of the cobwebs.
2 Likes
Tracks having the same AcoustID is strong evidence that they are from the same source material. Not guaranteed, I think only the first two minutes of sound is fingerprinted. But I can use the AcoustID field in MusicBee to help locate duplicates and decide to delete or not.
Over the last 25+ years, I ripped my CDs, first to VBR MP3, and later to FLAC. Of course I had an MP3 player that wanted FBR MP3, so I had to transcode to that. Then my car audio wanted something different with the tags. And the DLNA server on my NAS. And so on. I wound up with several copies spread over different devices. Now most of the time I just want to delete the lower quality versions that I don’t need anymore. AcoustID is another data point to verify I’ve identified the right files.
2 Likes
I feel your pain 
. I am “album-centric” and after many false starts settled on my present process. Rip once to “album with cuesheet”, embed MB releaseID in the cuesheet (keep backups). Now with just that I can recreate anything I need. I am starting my vinyl rip and will create a cuesheet for that rip so that it fits into my process. I do have a minimal amount of digital music I bought, still thinking about how I fit that into the present process.