Great! I haven’t fully understood the reasons to remove this ID as most of these multiple IDs are from a time predating NGS. But it’s a good thing and I will remove this release from my “Multiple DiscIDs” collection and cancel my annotation.
Also to add to the confusions of this thread, I’ve just been editing a Release where the edit history is showing a bored editor moved two other discIDs into the release without any explanation at all. It all gets very confusing as to the history of many discIDs.
It would be really handy if there was some way of leaving notes of verification comments against discIDs. Even an annotation field. Or a way to verify which disks people really have.
2 Likes
Hey ernstlx!
@Freso kind of beat me to it, but I really would just remove the DiscID’s. People are (very) unlikely to move DiscID’s about, while they are more likely to notice theirs is missing entirely and then add it (and can then be asked what version it is). Adding blank releases seems really confusing in comparison!
p.s. I assume this has been mentioned before but the same release can legitimately have multiple different DiscID’s
4 Likes
I did not expect people to do anything with this release or its DiscIDs. In fact I thought of having a disambiguation comment like “do not use this release, add your own release instead”. I see it as a method to keep these DiscIDs, but separated from well documented releases.
Really? I thought, different TOC = different release.
Yeah - this is very common. Extra pressings can be made with the exact same artwork being reused. All kinds of weird things go on like this which is where there is so much doubt about deleting discIDs.
The more popular an artist, the more likely it can happen. Especially if the disk is in production for many years.
Would be good if the Matrix details could be attached to a discID.
As I understand the guidelines it is more about finding a visual difference between the releases.
And today I bumped into an example of this. EMI SWINDON pressed on the matrix, but CD actually pressed on the EMI UDEN plant.
Look close at these two releases. Specifically the CD and Matrix photos. https://musicbrainz.org/release/2733229c-16f1-4700-85f7-ef5143da1802 and https://musicbrainz.org/release/36c26e35-2e90-48e8-9e0f-56cb6504c875
Notice the matrix on one has the extra -NL added? Even though they both say EMI SWINDON?
Now look at the fronts of the two CDs. One is Made in UK, the other Made in Holland. Same with the paperwork on the back.
This is an example of a Glass Master made in SWINDON and then sent to UDEN for manufacture. Whilst also copies are being printed in SWINDON at the same time. I expect in this case that both the Made in UK and Made in Holland disks will have an exact same discID. 
2 Likes
But there will be no distinct track times for this release.
This is an example for two releases sharing the same DiscID. And, of course, that you can’t trust anything that’s printed anywhere. 
Is there an example for the same release with multiple IDs? (legitimately added)
1 Like
I have my eyes peeled for that. And I am pretty sure I have seen it. I know that I have had trusted editors add discIDs to releases that I have also added a different discID to. Just can’t remember an example yet.
I am sure I have a disc somewhere with “Made in EU” on the paperwork. That then leaves it open for the CD to come from any pressing plant across the EU. Especially for those popular releases. All depends who then made the glass masters in those cases.
I’m still learning, but the deeper I did the more complex it gets. 
Leaves a lot of doubt when it is not clearly written on the disc or paperwork.
2 Likes
When I add matrix information to release annotations and I also have the CD physically, I usually make sure that the Disc ID gets listed together with the matrix data; e.g.:
(Note 2nd variant being sourced from Discogs data, so no access to get/check Disc ID.)
(Feel free to look through my release annotations for more examples for different edge cases relating to adding matrix information to the annotations. This is going off-topic though, so if you want to discuss this more, reply as a linked topic. )
6 Likes
Thanks. Nice template. I’m going to use that on the more common artists with multiple discIDs. It will help flag up the “correct” discs for other editors.
1 Like
Here is a short example of “Why would there be more than one discID”.
Pink Floyd UK Disc Manufacture
You may have noticed by now, I am a bit of a Pink Floyd nut. I have a few dozen copies of Dark Side of the Moon. All the same music, but the discID has changed dramatically between versions. Even number of tracks (early CDs merged first two tracks as one). In my case all my artwork is unique between versions so all of mine are separate MB Releases. But that is not always the case.
This CD has been fairly popular world wide and has been constantly re-pressed over the decades. That short description linked above walks you through the different manufacturing locations, explains why masters change, and all kinds of details. And this is just for one album!
Follow the link below for the matrix details on different versions of UK releases. The first Harvest edition being a good example of many variations.
http://www.pinkfloydarchives.com/DUKCDPF.htm#DSOTM1
I don’t yet have enough knowledge to point at where discIDs actually change in these lists, but you can see why there can be so many variants.
Errr… and this is just the UK up to 2004 before manufacture moved to UDEN… there are separate pages for Europe, USA, Japan… 
3 Likes
Do you add such matrix information to all of your edited CDs? Should we all do so? I only added scan images but nothing textual. It’s a pity that there no fields in MB for adding such information in standardized form.
Regarding variants:
Do you expect other variants to have a different TOC? They are manufactured at the same plant but on different moulding machines. Probably they originate from the same master.
So whenever artwork is the same, it’s the same release? Even if the track numbers are different?
Often I skipped existing releases because of other differences without having a single look at the artwork. (and very often there is none, and Discogs has mostly low res images and proved to be not very trustworthy)
I’ve noticed and that’s great, because there’s probably not much work left to do, when I will enter mine 
How would you know if any of the CDs within one release would have a different TOC. They total time given is the same for all of them. And the varying track numbers are from a separate release.
Do you posses more then one CD from one release?
2 Likes
Sorry if I am confusing you. Differences in artwork make an obvious point for a different release. Even the little differences on a rear cover with price codes make a different release. Different pressing plants, different release.
Differences in tracks lengths should also create a different release.
Certainly differences in number of tracks is easy to spot. And that is clearest way to say “new release”. (Even the non-nerds know “different number of tracks means new release” )
Errr… have you seen how many different Releases are listed at Discogs for something like Dark Side of the Moon? They have 963(!) compared with only a few hundred at MB.
There are always more versions to be found and added… it all depends when you purchased yours. Check out that pinkfloydarchives link I posted above. I expect you’ll need to click through to the European versions depending on your year of purchase.
Pink Floyd Archives-Austrian CD Discography <— Austria
Interesting as you also get a short “CDs manufactured in Austria” on the top of that page.
Pink Floyd Archives-E.U. CD Discography <— Europe
Again, scroll down the page and now the European history pops up for you…
I would not be surprised if you have yet another variation 
Errr… I don’t. This is the problem. Maybe the differences are only in fractions of section when a new glass master was made. I am still learning - I don’t know the exact point a discID would start to vary.
Multiple discIDs on a release should generally only be a few fractions of second apart. Maybe different gaps between tracks. Or tracks are lined up to different sectors. DSotM is probably a bad example here… more likely to find bigger differences in discIDs with cheap compilation Releases.
In my case my versions tend to be different editions of some form. A couple on Vinyl, Original CD, 2011 remastered, 20th Anniversary, 30th Anniversary, etc. Versions in boxsets. Then I have live bootlegs, “borrowed” copies… (I refuse to pay £100s for those new reissues boxes. I may be mad, but not that mad :D)
I don’t have any duplicates so can’t directly compare discIDs in that way.
-=-=-
One of the troubles is that not every editor who submits a discID is the Full Nerd 
. Not everyone cares for the tiny details. Some are less interested in attaching a discID to the correct edition and just attach theirs to the ones with the “prettiest artwork” so that Picard can download. This is especially likely when there are dozens of versions to select from.
You see this pattern when there are many extra discIDs attached to the “first” release in of a CD in a country. Or that Garbage example from earlier. You can spot this attitude with a few noobs passing through the forum where there are people just tagging thousands of files in one go. They don’t really care for the details, just want some artwork. These similar people will not worry that their CD is not exactly the correct edition when doing a CD lookup.
It is left to us detail obsessive nerds to untangle the wheat from the chaff. What @Freso is doing is a clear way of making it more obvious to other editors which discIDs can be trusted. I know I will be going back through all my Pink Floyd CDs to do that at some stage.
Personally, for Pink Floyd, I have also been adding in links to that PinkFloydArchive as that site is also clearly understanding the deeper nerdier rules of spotting the subtle differences.
2 Likes
But this would cause every CD with a different TOC to be a different release. So no multiple DiscIDs on one release. (?)
Probably not only variation
I try to limit the number of releases with the same content I don’t buy compilations, even of favored artists. (except there is an unreleased track included … it always is … )
And most inexperienced editors would know about this details. A short while ago I haven’t known too.
1 Like
DSoTM is the wrong example here. And where my knowledge starts to get shakier. The audio you hear could be the exactly same, but the gaps around it may differ. If all external things show that it looks the same - Packaging, CD image, Track Count - then I believe these would be the same release.
I believe a different mastering could have the tracks on the CD can start at a slightly different point on the physical media but still all be exactly the same length. The discID would be generated as different as those sectors numbers would be different even though the audio would be exactly the same.
Ripped to FLAC this would generate exactly the same audio, but the discIDs would be different due to the layout of the CD.
Or at least that is how I currently understand it. And am fairly sure I have an example of this somewhere in my collection but can’t remember where.
DSotM is (almost) the only album I have like this in my collection. But that is because the audio quality of the original CDs was junk compared with the remastered versions that came out later. Yes I can hear the difference in my anniversary versions.
With DSotM I also have a lot of bootlegs of concerts - some as CDs, some from slightly more dubious sources… 
I will not admit that here that when I first found Picard I too didn’t understand this stuff. I know I added discIDs using Picard to releases that “looked right” but now I know were totally the wrong releases. I can’t see me being the only one who has done that.
The good side of the way that @Freso adds the annotation notes is this is a proper nerdy way of doing things. Data like that can be double checked and trusted as the is less doubt. With a long list of discIDs there is doubt from both sides - just can’t be sure if they were added with care or “close enough”.
Pity that adding a discID isn’t like adding an AcoustID. That would then at least give a better weighting to the “good” additions.
1 Like
I don’t think so. Ripped tracks would include silent parts and thus would have a different length.
Variants on Discogs (those I’ve seen) are always manufactured at the same plant (IFPI NN**) at about the same time (if it’s for the same release), so they most probably derive from the same master. I would expect (at least for the 99% case) there will be no second valid DiscID for one release.
I’m the living proof that you’re not the only one. 
From my side definitely!
1 Like
What I meant is only clear by looking at the TOC in a discID or EAC log. Notice the sector times. What happens if the first track starts later on the physical CD, but all the gaps between tracks are then the same. That would lead to a different discID but all the tracks would still be identical.
For something like DSotM you can’t change those gaps between the tracks as all the tracks run into each other. But you could start the first track a little bit later on the physical sectors.
I think looking at those Peter Gabriel concert CDs would be a good alternate comparison. They will have been produced in a much lower production run. Making the CDs much less likely to every have generated a second discID. Personally they are the kind of disks I often skip scanning a matrix for as I doubt they will be reprinted.
In many ways I agree with you - stripping out the more obvious “wrong” discIDs would be good. But I trip over the “if in doubt…” rule because it is really hard to be sure. I have moved a few discIDs before, but never deleted any.
Some of the confusion can come from the lack of documentation. I’ve only learnt about the smaller differences via edits and forum discussions.
https://musicbrainz.org/doc/Release
It is good to go read at Discogs about this kind of stuff as they have some extremely detailed pages on the differences.
2 Likes
But this seems to be a really rare case. For all TOCs I’ve seen so far there’s always an exact 2s or 150 sectors track 1 (except with a hidden pregap tack). Look at the TOC of my example CD:
| Track | Start Time | Sectors | Length Time | Sectors | End Time | Sectors |
| 1 | 0:02 | 150 | 5:13 | 23460 | 5:15 | 23610 |
| 2 | 5:15 | 23610 | 6:28 | 29097 | 11:43 | 52707 |
| 3 | 11:43 | 52707 | 6:10 | 27753 | 17:53 | 80460 |
| 4 | 17:53 | 80460 | 3:56 | 17707 | 21:49 | 98167 |
| 5 | 21:49 | 98167 | 6:56 | 31203 | 28:45 | 129370 |
| 6 | 28:45 | 129370 | 8:10 | 36720 | 36:55 | 166090 |
| 7 | 36:55 | 166090 | 4:17 | 19292 | 41:12 | 185382 |
| 8 | 41:12 | 185382 | 5:58 | 26850 | 47:10 | 212232 |
| 9 | 47:10 | 212232 | 4:18 | 19368 | 51:28 | 231600 |
| 10 | 51:28 | 231600 | 10:11 | 45807 | 1:01:39 | 277407 |
There are no further gaps. For example track 3 (which is track 2 on the album, of course) starts at sector 23610 continuing from 2 and ends at 52707 resulting in a duration of 29097 sectors = 6 minutes and 27.960 seconds (27 + 72/75 seconds). There's not the faintest gap not belonging to one of the tracks and that's true for all TOCs I've seen so far. (again except with a hidden pregap tack, but even there are 2 additional seconds compared to the content of this index 0 track)
If you can provide an example where this would be different, please do so. Anyway, I suppose the vast majority of albums will have this pattern. So probably most of the releases will have only one valid DiscID, don’t you think so?
I haven’t done this so far. I’m not even registered, but I will. It’s no use complaining about all kinds of errors on Discogs…
BTW, it took me quite a while to learn of the limited HTML capabilities of this forum, and some strange phenomena appear in the preview still… 
1 Like
I try to.
No. People have different levels of need towards adding verbose data and people have different amounts of time and energy they’re able and willing to spend on the project and have different goals in this regard. I choose to spend my time on energy in “combing” the CDs I add for SID codes and matrix info and adding that to the annotation, but I do not at all expect other volunteer editors to do the same.
I would expect them to have the same TOC (and thus Disc ID), but the only way to track whether this expectation/hypothesis is correct or not… is to actually track it.
8 Likes
Of course, I did not expect all editors to be able to do so, but I decided to add this information to my releases. like you did as I determine it anyway.
I will track it, at least for releases I’ve added to my collections.
A bit off topic is another album I’ve added recently. Although it has 2 DiscIDs (like half of the other 17 releases), that’s not the main problem. It might well be that this release consists of up to 3 different releases. Only a look on some artwork details could help to decide if that’s the case or these are variants of the same release. I know only about mine, but @Fabe56 originally added a possible French release, which was later extended to be FR+GB release to match the linked Discogs release (only GB). I thought I have a German release, but there are many similarities with the existing release(s).
(“special limited edition” or something like that in another language)
I do not add details here as I wrote an extensive edit note and probably only people holding the physical release in hands can solve this problem.
3 Likes
Luckily I’m in France, with access to this part of my collection. I’m gonna go check it out
3 Likes