Works with many different segmentations

MusicBrainz Classical
, ,
1

Hi!
I’m currently trying to clean up Schütz, and I have run into a problem several times:
Some of his works, especially the oratorios, do not have a default segmentation like others, there are no clear movements so everyone splits it his own way:
As an example, the split of the Johannes-Passion on
https://musicbrainz.org/release/0d3ff72f-0c20-4d0c-badf-b88ca2d5058c and
https://musicbrainz.org/release/12ec71e4-341d-4ab6-85ea-f217e2b2b924
Now I’m wondering how to handle this correctly: should I split the work into its tiniest possible bits and then each recording will have several works per track (I did this with the Matthäus-Passion https://musicbrainz.org/work/363d7924-59c2-413b-a2bf-638ab0584a89 and ended up with parts that were about 3 seconds long sometimes) or should I just keep the whole work and make only partial recordings of anything? The first seems better on the pure data standpoint, but it is a real pain to enforce and maintain. How do the Wagner specialists deal with it? I noticed this on the Wagner operas as well.

Also, I noticed that on Schütz (and I guess on others) there is a tendency to replace the composer by the performers in the recording artists before a work is added. This makes it impossible to do what I’m trying to do (go to the Schütz page, go to the recordings tab and process everything that is there). Is there a way to clean up a composer that does also spot recordings where the composer has been removed as a recording artist?
If not, is there a way to, when I go through the latest edits for Schütz, to spot the recordings where the composer was replaced as a recording artist without a work being added (maybe an userscript) without going to the page of each recording to check?
Thanks,
Dorian

1 Like
2

Generally, for cases like these, I just use partial for the whole work. For example, Liszt’s Piano Sonata.

3

I’m wondering if it still makes sense for works of the scale of an oratorio. It seems that the opening and closing choirs are consensual and well identified, maybe I make a work for the opening, one for the closing and one for the bulk.

4

I did this with a couple of operas, and it is, as you say, a real pain.

That seems a sensible approach: choose the divisions that are pretty much universally acknowledged, and don’t try to subdivide further.

It’s still probably inevitable that different recordings, and different editors, will choose to divide the work into different sections that overlap. This is unfortunate from the point of view of the work entity, which won’t have a clean division into sections and subsections, but I don’t see an easy solution.

5

Funny, I thought of an OP on this, triggered by Schütz’ Matthäus-Passion, and it was here already!

Just added a 1989 Release

in same Group as the 1991

Obviously the same parent recording cued to 7 tracks in the one release, 58 tracks in the other, and none shared. Of the tracks betitled “Barrabam”, 1989 #36 is 11 seconds (chorus only) while 1991 #5 is 1 minute 30 seconds. Another release 2000 (2CD)

has the same recording in 7 track layout with 1m30s cut of Barrabam as the 1991 release, except for 21 seconds “moved” from 2nd track to 1st (here: 2.4 → 2.3).

Yes. The work has 3 formal parts: Introitus (1) – Evangelium (40) – Beschluß (4)
Number of rough score pages in parethesis to give a hint on proportion.

The MB work entry had several parts, not all main 3 though and only 3—4 of the ~60, in a roman numbering that more likely comes from a release tracklist than a score; not all set in specific ordering, and all on the same ‘part of’ relationship niveau. I have done some lossless tidying:

Edit #59044147 had the bunch reordered.
Edit #59044229—31 had the 3 subworks removed as parts of Passion, and after some having had them re-inserted as subworks (parts of) the Evangelium, the new layout will be like:

Schütz: Matthäus-Passion

  1. Matthäus-Passion, SWV 479: Introitus: Das Leiden unseres Herrn
  2. Matthäus-Passion, SWV 479: Evangelium
    2.1 Matthäus-Passion, SWV 479: XXII. Und der Hohenpriester stund auf (Evangelist, Kaiphas, Jesus)
    2.2 Matthäus-Passion SWV 479: LIX. Herr, wir haben gedacht (Hohepriester und Pharisäer)
    2.3. Matthäus-Passion, SWV 479: LXI. Ehre sei dir, Christe
  3. Matthäus-Passion, SWV 479: Beschluß

Aware of “deepening” the Work this way, I find it better than the previous mess; open for completion, still with a “catch-all” Evangelium that is cued to many different Track layouts on releases even of the same parent recording. Are there objections to such ‘multi-layering’ of works’ ‘part of’ relationships?

Another work with recurring different track mapping is Bartók’s Str4 1 where the Introduction to III. Finale is often not “prefixed” to that track, but instead “suffixed” to the II. track. The score is headed: I. – II. – Introduction – III., clear on artistic intent.

6

Yes, I remember merging a lot of subworks and just keeping the subworks everyone seemed to agree on (mainly the choruses, entry and conclusion). Obviously I didn’t get completely through ordering it, thanks for finishing the job! I have a remark: It seems like the last two works are the same (https://musicbrainz.org/work/b3077f9f-231c-46b3-a2a9-0a00e98a0152 and https://musicbrainz.org/work/32ffe837-92b4-4188-9b8c-08e127cc5ebf ). Will merge them.

7

Of course, I ran through not checking incipit on that one since it was missing its “Beschluß” while “Das Leiden unseres Herrn” had its “Introitus”. I see you have already sorted out back the ‘part of’ relationships too. Thanks.

Edit: I should not cancel #59044231 of these 3 edits, should I?

8

Actually no, else it risks having the relationship twice. (edit: I was in favor, but it turned out I checked the wrong edit)

9

@reosarevok I’m wondering if you have an opinion on this: https://musicbrainz.org/edit/59043919 . It seems to be that we have two times the same recording, but with different cuts (one cuts ~20 seconds earlier, so the first track is longer and the second shorter). How do we deal with this? For me the recordings should not be merged, but I can’t find an appropriate relationship.

10

Just to clarify to others: The edit concerns 2 tracks of the same recorded interval, that on 2 releases have been cut diffently, but our crux has ramifications: It is about the releases in the above post. The 1989 one has the same ‘parent’ recording cued to 58 tracks, none of which matching any of the 7 on the 2 (1991, 2000) releases in question. Here be dragons.

11

I am using the compilation relationship in similar cases, e.g. https://musicbrainz.org/recording/fc0631a9-2d8b-4eb6-9a16-55066b53352d (following reo’s advice)