Importing textual data across albums

brainchild0 · June 3, 2021, 7:25pm

I own the album photographed below, a recording of the Brandenburg Concertos.

The cue sheets also are shown below.

They resolve to the following two disc identifiers:

Currently, only the first has an entry in MusicBrainz, even though both belong to the same album.

I’m not particularly clever at entering all the proper details into the system, as someone has done for the first disc, but I notice that other items in MusicBrainz have essentially the same information, at least with respect to textual metadata despite very small differences in the exact track layout.

For example, when compared to the metadata for the disc referenced above, album be23c3cd-9bad-48ab-824f-a717e214a5d3 is shown as essentially the same. As seen too, the slight differences occur in the layout.

I am not fully educated in the relationship and uses of releases, disc identifiers, and other item types. I am also unable to consider any reason why two albums would be so similar but yet slightly different.

Is any preferred method available into import the metadata from the album with a complete representation to the one with the incomplete representation, assuming that the correct information for the missing fields is indeed the same?

Cue sheet for first disc:

FILE "abcde.6f09ba09/track1.flac" FLAC
  TRACK 01 AUDIO
    INDEX 00 00:00:00
    INDEX 01 00:00:33
  TRACK 02 AUDIO
    INDEX 01 04:04:65
  TRACK 03 AUDIO
    INDEX 01 07:49:43
  TRACK 04 AUDIO
    INDEX 01 12:02:55
  TRACK 05 AUDIO
    INDEX 01 19:38:10
  TRACK 06 AUDIO
    INDEX 01 24:44:35
  TRACK 07 AUDIO
    INDEX 01 28:30:18
  TRACK 08 AUDIO
    INDEX 01 31:17:13
  TRACK 09 AUDIO
    INDEX 01 36:50:33
REM FLAC__lead-in 88200
REM FLAC__lead-out 170 109844280

Cue sheet for second disc:

FILE "abcde.880c1d09/track1.flac" FLAC
  TRACK 01 AUDIO
    INDEX 00 00:00:00
    INDEX 01 00:00:32
  TRACK 02 AUDIO
    INDEX 01 07:21:42
  TRACK 03 AUDIO
    INDEX 01 11:02:45
  TRACK 04 AUDIO
    INDEX 01 15:34:32
  TRACK 05 AUDIO
    INDEX 01 24:56:22
  TRACK 06 AUDIO
    INDEX 01 31:06:37
  TRACK 07 AUDIO
    INDEX 01 36:18:72
  TRACK 08 AUDIO
    INDEX 01 41:43:05
  TRACK 09 AUDIO
    INDEX 01 46:13:30
REM FLAC__lead-in 88200
REM FLAC__lead-out 170 136775856

aerozol · June 3, 2021, 11:11pm

You can easily clone the tracklist and recordings across, which is 95% of the work

Adding the fields like label, barcode, cat number etc you should do manually, but that’s the fun part!

Duplicating the tracklist can be done in the MB interface, just start from the [release group]:

(^ click to view gif walking through it)

outsidecontext · June 4, 2021, 6:03am

Not sure I understand. So you are saying the release you have is a bit different from the one MB has in regards to cover art? Could you show an example?

If it is indeed different a new release should be added. If it is fully identical you should instead submit the missing disc ID to the second disc of the existing release. Although it is a bit odd that your disc has offsets of 2-3 seconds for each track. But the track length itself seem to be the same as far as I have checked.

jesus2099 · June 4, 2021, 7:18am

I searched Nancy Kammerer (credited for Product Design) and it leads to Nancy Kammerer | Discography | Discogs then to Musical Heritage Society Label | Releases | Discogs .

Your edition looks completely different than the https://musicbrainz.org/release/be23c3cd-9bad-48ab-824f-a717e214a5d3 and according to Nancy Kammerer, it’s an Musical Heritage Society (MHS) kind of mail-order-only CD.

So yes, you should create a new release in the same release group (if same content).

There is even another release with just the CD1, maybe they just did not submit their CD2: Release “The Brandenburg Concertos - Orchestra of the Age of Enlightenment” by Johann Sebastian Bach - MusicBrainz

outsidecontext · June 4, 2021, 7:53am

That could also explain the differences in disc ID. Quite possible they did repressings of releases that were still on demand, without necessarily considering this a re-release.

brainchild0 · June 5, 2021, 8:53am

A variety of details have been mentioned, including some, such as cover art, which are beyond what I was considering when I posted the original inquiry.

Broadly the question is how best to do the following:

Determine which data already entered into MusicBrainz, such as that in the other album, correctly represents the data for my album.
Transfer specifically that data into the record for my album.
Provide whatever further information then remains missing for my album.

For doing so, it would be helpful for me to understand why a different album would occur with essentially the same data but a track layout that is different only by the slightest amount. Do the small differences in track layout imply the need to be wary of reusing the same text?

foxgrrl · June 5, 2021, 8:59am

Does the ring around the central hub say:

724356155329 MASTERED BY EMI MFG. 1-1-1

or

A 7301 724356155329 E50125-14

foxgrrl · June 5, 2021, 9:14am

So… MusicBrainz currently doesn’t keep track of Matrix codes, but it does keep track of CD cue sheets. These two pieces of data tend to be tied together, so that when one changes, so does the other. But, MusicBrainz doesn’t distinguish between different masterings, engravings and pressings of the physical CD itself, only the printed packaging – silkscreen printing on the CD itself, and all the paper stuffed in the jewel case.

So… the standard procedure is just to attach all the CD TOCs to whichever release the packaging matches… but, what I’ve been doing personally, if I can’t confirm that my release is in fact the exact, exact, same one as something already in the MusicBrains DB. I’ll add a new release, and record all of the details about the Matrix code and IFPI codes, and stamper code engraved on the hub, etc.

Because it can always be merged into another release someday, but if in the future, MB starts tracking Matrix codes, it’ll be a pain in the ass to separate all the DiscID back out again.

brainchild0 · June 5, 2021, 10:28am

@foxgrrl: I find neither sequence on the inner rings. Instead, I see a sequence that also appears on the front of both discs, followed in either case by 01 or 02, to indicate the disc number.

The two sequences all together are MHS 522498 01 and MHS 522498 02.

Regrettably, I’m not familiar with the more technical terms in your explanation. I am just a casual consumer of music.

My confusion is that whereas variance in timing might be natural for separate recordings of a composition, and redesigning the artwork might be natural for later publications of an earlier recording, small variations in timing for different publications of the same recording are difficult for me to try to explain.

brainchild0 · June 5, 2021, 10:51am

The following appears in an explanation of Disc ID:

Different pressing of a CD often have slightly different frame offsets, and hence different disc IDs.

Perhaps the situation I have described is common, but it surprises me, for reasons given.

foxgrrl · June 5, 2021, 10:56am

Once upon a time, before large capacity hard drives existed, recording studios would send CD pressing plants the digital audio data for a CD on video tapes. At the pressing plant, somebody would manually type the track offsets for the TOC into the laser engraving machine, which zapped a smooth polished glass plate full of tiny holes. This was coated with an extremely thin layer of metal, and used as a mold to cast a stamper, which is then put in another machine and squished repeatedly against blobs of polycarbonate plastic. Eventually this stamping would wear out, so they’d make another one… and another one, if it was a really popular CD.

Anyway, after the plant pressed a bunch of CDs… a year or so later, if the record company wanted some more, they would need to start this whole process over again (if the plant didn’t save the original engraving, which they usually didn’t… or the record company used a different pressing plant) Sometimes… you wouldn’t need to start all the way over with the CDDA data on a video tape. You might just start by reading the CDDA from one of the CDs in the first pressing… and maybe, just maybe, you would get the same CDDA data as the original… but usually not. So that’s why there’s different pressings of popular CDs, which only differ by a slight offset, or the occasional transient impulse “click” error on some track.

foxgrrl · June 5, 2021, 11:19am

The Compact Disc was designed in the mid-1970’s before microcomputers were really a thing. Other than the particular encoding they use, they are not in any modern sense of the word, “digital”. Think of the CD as a small vinyl record, with a 2.5KHz analog FM signal recorded on it. Because that’s how the designers though about it, and it’s exactly how the CD actually works. You run this signal through a 14-bit analog to digital converter, and the error correction decoder used for satellite broadcasts, and ta-da, you’ve got your high fidelity sound wave back.

The table of contents (TOC) is an optional afterthought. CD Players can ignore it entirely, as some early CD players did. You can drop the needle, I mean laser, at any point on the CD and just start playing from there. There is no absolute addressing, no sectors, no identifying position info except for the minute, second, and 1/75th of a second frame count encoded in the “Q-subcode” channel… which is spread out over several thousand bytes of PCM audio data.

Anyway… in the official Phillips CDDA Red Book standard, the timing (position) of anything doesn’t need more than two seconds of precision. That include where the audio program actually “begins”. So every CD engraving machine would start engraving at a slightly different offset after the end of the lead-in zone (the TOC and stuff).

So, that’s why the offsets move around by 700 samples, or whatever, between releases.

aerozol · June 5, 2021, 11:39am

This discussion is getting pretty complex, so just to be clear:
If the images you posted are of your version, and the DiscID matched to this release, then you need to follow the steps that I described, because different cover/packaging = new release, always.

The discussion around track timings, TOC’s, matrixes etc (particularly what @foxgrrl posted, very good reading) is fascinating but it doesn’t really impact how you add the release FYI

brainchild0 · June 5, 2021, 12:16pm

I just realized that both parts of the release have two associated Disc IDs, but a match occurs to my Disc IDs only for the first disc, whereas Disc ID of my second disc is not known in the system, even though other Disc IDs are already associated.

Is this situation typical or unexpected?

brainchild0 · June 5, 2021, 12:25pm

Also, for some Disc IDs associated with the release, an option is available called “Set track durations”, which shows a particular set of adjustments.

I am confused. I understood that a Disc ID is computed deterministically from the track layout. Also, how is the system determining the adjustments?

foxgrrl · June 5, 2021, 12:34pm

This sets the time durations on this particular release’s (the one it’s attached to) tracklist. And then makes the times un-editable in the track editor (they’ll be greyed-out). It might update the recording times too… maybe…

foxgrrl · June 5, 2021, 12:39pm

I’ve seen this happen before… I’m guessing that years ago, someone only added the first Disc ID, and just didn’t add the second.

I think the only two sources of Disc IDs are: 1. The CD Baby catalog and 2. User submissions from Picard
oh, of possibly also 3. The FreeCDDB or last GPL’d CDDB databases, which were a mess.

outsidecontext · June 6, 2021, 6:02am

As part of the disc ID also this track layout, the data from the TOC gets submitted. This includes the offsets of all tracks. The data allows setting track durations that match that track layout exactly.

brainchild0 · June 6, 2021, 1:12pm

So the user making a submission has the option of aligning the release track durations to the table of some disc, versus entering them manually. This point still confuses me deeply. How can the same release correspond to multiple discs with different tables? Is it just because of the technical issues making presses match identically that leads to different discs being created from the same release?

outsidecontext · June 6, 2021, 1:43pm

I think foxgrrl has some excellent explanation above why slight differences in TOC can happen in the production process.

In my experience older CD released have much more different disc IDs than newer once. Probably due to the manual work involved for older CD pressings, where not everything of the process was digitalised.