Recording split in 2 in later release: which relationship to use?

Comparing two releases of “Eugene Onegin”: first and second. The same performers (names available on CD covers), many tracks of the same length. But some tracks from the first release are split in the second release on 2 or more recordings. For example, the very first track from the first release (duration 7:05) is split in the second release on two tracks (durations 2:04 and 4:58).

Looking on discogs entry for the first release, it appears to be the original, since the first versions of it were release as long ago as in 1961.

What type of relationship between recordings would you use in a case of a split, as in the example with track #1 from the first release and tracks #1 and #2 from the second release? “Edit of”, “remaster of”, “later releases”?

1 Like

I’d use the Recording-Recording / Compilation relationship.

This indicates that a recording is a compilation of several other recordings. This applies to one long recording that contains multiple songs, one after the other, in which the audio material of the original recordings has not been altered. If the tracks are pitched or blended into each other, the DJ-mix relationship type may be more appropriate.

1 Like

English is not my first language, but doesn’t “compilation” imply that original version had many short recordings, and later one longer recording (compilation) was created by merging original recordings in one? In my case it is the other way around: the original recording was long, but in later releases it was split on multiple short recordings.

There’s a bit of that implication, but we don’t really differentiate between those two cases. The main point is that if you have the “big” recording, you also have all the “small” recordings inside it.

1 Like

That (pretty clear) meaning is another instance of non-common language use to MusicBrainz.
Each instance of non-common usage steepens the learning curve for new users.

Some instances of non-common usage by MusicBrainz appear to have nett benefits.
This instance could well be one.
(The costs of non-common usages can be reduced if the Documentation makes clear that MB is using the word/s in a non-common way.)

1 Like

One more case, similar to the previous, but slightly more tricky.

First release: full recording of “Eugene Onegin”.

Second release: highlights (excerpts) from “Eugene Onegin”.

Same performers, same conductor, same label (Deutsche Grammophon). Many recordings has practically the same length, difference 1 second. I can’t find a scan of a back cover of the “highlights” release, but everything (incl. identical length of corresponding tracks) indicates that “highlights” were created from the first (full) release.

Now the tricky part. Full release, disk #2, track #11 - length 14:15. “Highlights” release, track #13 - length 11:32. If one knows the content of the opera and looks on track names, it is obvious that #13 from “Highlights” is a part of #11 from the full release. Would you consider all the circumstantial evidence enough to prove that “Highlights #13” is indeed a part of “Full #11”? If yes, what relationship should be used between recordings?

1 Like

I’d be hesitant to just use the circumstantial evidence - but I think other Editors here would find it sufficient.

Is there harder evidence available?
Can you listen to the two Recordings?
Does AccousticID offer a viable way of comparing these Recordings?

“Edit of” is what I’d suggest here (if you’re confident enough they’re the same).