Could somebody explain what the exact conditions are for Picard to write [non-album tracks] for a recording?
My reason for asking (without going into too much detail) is that it is writing that sometimes in situations where I would rather have Picard not doing that.
edit,
I can now probably answer my own question.
It probably only happens when you have ‘Ignore MBIDs when loading new files’ unchecked, and the file does not contain a musicbrainz_albumid tag.
It does this if you load a single recording instead of an album into Picard.
But not if you have ‘Ignore MBIDs when loading new files’ checked, and then scan, and have the track matched, correct?
When “Ignore MBIDs when loading new files” is set Picard will not load any release data for files with existing MBIDs. If this is unchecked Picard will load a recording without album if the file as a recording ID but not release ID.
Btw, there is a ticket to improve this a bit so Picard would also consider a release group ID (which it currently completely ignores). The idea would be that if there is no release ID but only a recording ID + release group ID Picard would still load the data and use it’s usualy matching to select the best fitting release.