The "When is it a soundtrack / when is it classical?" question

Tags: #<Tag:0x00007f0509aef708> #<Tag:0x00007f0509aef500> #<Tag:0x00007f0509aef348> #<Tag:0x00007f0509aef168>


Has MB got librarians on board?
They are trained and experienced in cataloguing. And sure they’ll push their own view of which is best but they are , or at least should be, very aware of the different approaches and the strengths, weakness, cost and benefits of the various approaches.


And any cataloging we choose to do should not impact the basic information like album name et al.
We should use the meta data for that kind of purpose, that:so all I mean.
So yes I probably mis-interpret in the first place. :cat:


Well then we agree anyway :stuck_out_tongue:

Artist intent doesn’t apply to completely internal attributes like disambigs, there MB is the ‘artist’ of our system and we should try to be consistent with our ‘intent’! As always, I still agree with your overall sentiment of ‘as on packaging’, and hopefully MB supports it better soon.


Cool, and BTW outside of classical that I don’t really edit, I feel that soon is yesterday, things have improved already a lot with AC, AC on AR, unlinking track texts from recording texts, etc. :sunglasses:


I think all the “specific types of releases” (STRs) are treated differently from a random release because they aren’t random :slight_smile: They all come out of somewhat different traditions, and MB tries to reflect that. STR guidelines are there to override the as-on-cover data in specific contexts.

If I was entering a random release that mixed orchestral music with pop, I’d enter it using the standard guidelines. The as-on-cover credits are probably fine. It’s the context of the release that matters.

Is it possible to unambiguously determine the context of every release? No. But that’s okay.


Related to this is how do others know what guidelines you are following. For example the guidelines for classical releases are quite different to the vanilla Rock/Pop standards for (track/recording) artists but how do you know which guidelines the editor has used. I feel there should be at least be some flag saying this release is considered classical has been entered according to the classical guidelines as opposed to a classical release that has been added using the vanilla guidelines. Without such a flag how is the data to be interpreted, especially if the interpretation is being done by computer rather than human.

Composer Tag

Off-topic for the discussion at large, but I just wanted to chip in that @Bitmap is working on this for the coming schema change release. Follow this for progress:


I haven’t read this topic (and sorry for the additional off-topic-ness), but note that that ticket is only intended to support alternative transl*ations, so if there’s an additional use case being discussed here that you think is important to support, I would comment on the ticket soon. :slight_smile:


This idea has the great benefit (if implemented) of allowing all “classical” releases to be IDed in the database, which would then allow a user to use strings or whatever to reformat metadata to their preferences - this would allow all users to be happy with the classical metadata provided by MB. (As far as I can see.)


Pretty sure that’s already possible with performance ARs, work/recording names, etc.


I’m swimming but as I understand things, at present the user who wishes to reformat the classical metadata has to select each classical release manually.
If there was a classical flag then they could select all their classical releases with one click.

This ‘flag’ approach could also be used as a stop-gap solution to the classical/soundtrack quandary.
Create an ‘edge case’ flag and if a release is disputed wrt classical/soundtrack then after very brief discussion, flag it as an edge case and let the active editor continue unimpeeded with their good work.

We would effectively pass the classical/soundtrack problem onto future generations of editors. And surely they’ll be more intelligent than us and come up with a simple fix.


Classical is not really my thing but it makes complete sense to somehow tag if we use classical tracklist style or not.


Can I get a link to one of these disputed releases?

If you go by the nature of the release, it seems clear to me. Something is packaged/presented as a soundtrack or it isn’t.


My assumption here is that this should be a workflow thing in the tagger: If I wish to tag a particular release in a particular way, it’s up to me to configure my tagger to use Musicbrainz data appropriately. For example, “$set(artist,%composer%)” (If I recall Picard syntax correctly).

If the data doesn’t exist in Musicbrainz, I can add it; if Picard doesn’t know about a specific AR I want, then it can always be changed so that it does (or I could use another tagger which does understand those ARs)


The problem is if I dont know what guidelines are being used I don’t know what the data represents. So for example the track artist doesnt usually store the composer (unless they are also the performer) for Pop/Rock but it generally would if following Classical guidleines. The trouble is many serious classical users really do not want the composer to be set
as the artist (.e Retrieve Album/CD title using Track Title, Track Artist, Label and Catalog#)

so the Hawkes Picard example is the opposite of what is wanted, what they want to do is ‘not set trackartist if from Musicbrainz TrackArtist field if using Classical Guidelines’ which is why a flag would be useful

What is classical?

But you do: That’s what ARs do.

My point is that the person doing the tagging knows what they want to do, and can script accordingly. Not that the example will work perfectly for what everyone wants to do. Maybe someone wants "$set(artist,$if2(%performer%,%composer%,)). Maybe it’s "$set(artist, %performer[orchestra]%).

It doesn’t really matter what precisely someone wants, they should be able to script it into whatever they want without changing the track/release artists in Musicbrainz proper. It doesn’t even matter whether Picard is currently capable of getting the exact effect someone wants (e.g. there is nothing like “$recordingartist” in Picard). The principle is that the way someone wants to use the data is independent of what guidelines should say about how Musicbrainz itself wants to display that data.

Picard can always be improved; other classical-specific tagger utilities can be developed, but the data is all there in the ARs if you want to use it.


I agree with the ‘The principle is that the way someone wants to use the data is
independent of what guidelines should say about how Musicbrainz itself
wants to display that data’

but it should be possible for someorne to get the data from MusicBrainz and understand what it represents.

If the ARs have been entered they can be used as they are applied consistently. But they are optional and if they have not been entered they cannot be used. In contrast albumartist/trackartist/artist will always be available but it is currently not clear for a given release what the value of these artist fields represents as they are overloaded for different guidelines and there is no indication of what guideline is being used for a particular release.


I think that points more towards adding ARs where possible than it does toward redefining what a track artist means. I don’t know that it’s explicitly defined anywhere in Musicbrainz but I believe the general concept for the “normal” case is to get fairly close to whatever’s on the cover; a lot of what the guidelines do follows that. It seems to me that they are mostly there to clarify what you should do in the odd cases where that’s not obvious.

I think that’s the case with all of the “specific types of release” guidelines, as classical which has both performers and composers that apply to every track and release, and with soundtracks which sometimes don’t explicitly list any artists at all, and with stage musicals which prominently credit lyricists as well, and are often performed by groups that are unspecified and we’d rather not have in Musicbrainz as groups.


Hmm you dont seem to be understanding my point when you say ‘then it does towards redefining what a track artist means’ and ‘what you should do in the odd cases where its not that obvious’

I’m not for one minute suggesting we should redefine what it means rather that the Classical Release guidelines already do dramatically redefine what they mean without there being anything in the release to indicate that they have been redefined. And the whole Classical genre can hardly be described as an odd case.


The issue is that track artist means absolutely nothing. Generally in MusicBrainz, they kinda-sorta mean “who is given credit for this”. For classical, they still mean “who is given credit for this” (but they pick one of the credits instead of other). The difference is much smaller than it seems. There’s nothing anywhere in MB that says track artist should be the/a performer for non-classical: they’re usually who’s credited, but it can also be a producer / writer / narrator / remixer / DJ-mixer / compiler (I’m sure there are edge cases that are even less expected than that).

And that is, in my opinion, a problem rather than a good thing: the current MB structure forces an album and track artist even in places where it makes absolutely no sense, which is what causes struggles with stuff like classical in the first place, and it makes relationships feel like secondary data: in fact, relationships are the most important, useful data, and not only for classical music.