Would removal of a click/clatter count as a separate recording?

MusicBrainz
, ,
1

I have been editing the following release: Release “A Brief History” by Penguin Cafe Orchestra - MusicBrainz

This is a hybrid SACD compilation of the Penguin Cafe Orchestra.

There were a number of duplicate recordings. I have merged the ones that I have listened to and compared and made sure they were exactly the same. (I have all their albums so this was straightforward.)

Nonetheless, there is one that I am not sure I can merge. This is Rosalosis. The sole difference between the original album and the compilation is they cut off a brief clatter/click at the start of the recording, possibly removed a few seconds of silence from the end.

You can listen to the original recording here: https://www.youtube.com/watch?v=xyLj7ZlZ48g and the compilation here: https://www.youtube.com/watch?v=IEnNvNFlVJY

As far as I am concerned this is not musical. It is the same as an applause or a click/pop introduced by a tape/LP. It sounds like someone dropped a coin in a drum or something like that.

So, to this extent I would merge the two recordings, but it’s unclear to me if this is ‘correct’. I would be interested in hearing some thoughts on this one.

2 Likes
2

I would further add that the acoustIDs actually match for these two recordings:
https://musicbrainz.org/recording/92dad508-024e-43d9-a75c-95d7762c5b29/fingerprints

https://musicbrainz.org/recording/3e0cf6a6-0fc4-47ed-b261-e7f13c62d6b6/fingerprints

It is unclear to me if this is a mistake, or the manner in which acoustIDs are computed (with regards to clicks and pops etc).

3

An AcoustID is a group of fingerprints. The AcoustID server compares fingerprints and decides if they are “similar”. I am not sure which AcoustID you are referring to, but both hold multiple different fingerprints.

Here’s an egregious example where 2 entirely different recordings share an AcoustID (click “back to fingerprint”, then “back to track”, on each fingerprint):
https://acoustid.org/fingerprint/12635077/compare/14351739

If you change the offset to -50 (negative fifty), you can begin to guess why the AcoustID server came to the conclusion that the fingerprints are “similar”.