How to handle metadata-'empty' releases with one or more attached discids?

MusicBrainz Style
, ,
1

Hello together,

so sometimes you stumble upon releases with only a tracklist added but also with at least one disc id.

to me these releases seem more id-storage than anything we could use as a base for valid metadata.

specificly i encountered the phenomenon again at:

and it is a bit difficult in this case to do something with it, as there exist multiple issues we do not list and also we do not have a clue, to which ones these should be attached. also the editing history does not help.

now i think we have some options:

  • wait until discid gets attached to a different release with higher data quality, then remove discid until release does not contain any more discids
  • do nothing
  • set data quality to low
  • delete release
  • transfer discids to other releases (is there a guideline?)
  • merge with the MB-equivalent of the most ‘famous’ release on discogs (Discogs Statistics highest number of users that mark ownership)
  • copy metadata from a discogs-release that is not on related to MB yet to this ‘emtpy’ release

what to do? :slight_smile:

1 Like
2

These situations are never simple, black-and-white, clean-cut - they all need to be assessed on a case-by-case basis :slight_smile:

1 Like
3

Given the lack of data this might have been a freedb autoimport from back in the stone age. There’s no good option really with them - unless you can figure out from gnudb or somewhere what was the source, it’s probably best to just turn it into something that seems legitimate. I honestly would even be fine with just removing the discID if there’s no info at all to match it to a specific release…

4 Likes
4

DiscID nerd speaks up…

Check the edit history and you find there is no record of them being added to this release. Nor anyone ever using them to Set Track Lengths.

These are from the pre-NGS era when some point in time all discIDs were added to all Releases leaving things confused and unknown.

Compare the two discIDs and one is clearly a home brew. That CD with a longer length has every track as plus two seconds over the short one. A classic effect of a home burnt CD.

If I saw this I would delete that longer DiscID and see if anyone complains.

I also agree that there is no harm in deleting both discIDs and just trying to make this something cleaner from knowledge at hand. Better to have known good data that a combo of unknown \ unsubstantiated data.

6 Likes
5

For the record, what happened exactly is that pre-NGS, a single release was all the editions with similar tracklist.

One release with 3 release events, for example:

  • Pre-NGS Release A
    • Catalogue number ABC-123 bar code 123, release in France on 1977-06-24
    • Catalogue number DEF-456 bar code 456, release in UK on 1977
    • Catalogue number GHI-123 bar code 789, release in Germany on 1978-01-01
    • Disc ID a65eaz71468zae11=
    • Disc ID 3654zea354eza68a=
    • Disc ID zq6za817e6687azz=

Then the NGS migration lead to one release per pre-NGS release event and assign all disc ID to each.

Now 3 releases (from those 3 release events):

  • Pre-NGS Release A1
    • Catalogue number ABC-123 bar code 123, release in France on 1977-06-24
    • Disc ID a65eaz71468zae11=
    • Disc ID 3654zea354eza68a=
    • Disc ID zq6za817e6687azz=
  • Pre-NGS Release A2
    • Catalogue number DEF-456 bar code 456, release in UK on 1977
    • Disc ID a65eaz71468zae11=
    • Disc ID 3654zea354eza68a=
    • Disc ID zq6za817e6687azz=
  • Pre-NGS Release A3
    • Catalogue number GHI-123 bar code 789, release in Germany on 1978-01-01
    • Disc ID a65eaz71468zae11=
    • Disc ID 3654zea354eza68a=
    • Disc ID zq6za817e6687azz=
3 Likes
6

If this catch-all release is already in someone’s collection, I usually add my well identified edition with its unique disc ID and leave that catch-all release as is.

1 Like