Luckily I’m in France, with access to this part of my collection. I’m gonna go check it out 
3 Likes
When I find it again, I will do. It was just that initial sector offset that had change, and therefore knocking all the others off. The track lengths and gaps between were still all the same.
This is good to read. For a while I was worrying I was getting too nerdy 
You can still read their detailed howto guides without being a member.
Starting with the simple “whats on the CD and rear cover”: https://support.discogs.com/hc/en-us/articles/360004051833-How-To-Find-Information-On-A-CD
And getting deeper…
https://support.discogs.com/hc/en-us/sections/360001058453-Full-Submission-Guidelines
Example - a page on details about SID codes and those other items on the matrix and back of CD. https://support.discogs.com/hc/en-us/articles/360005054893-Database-Guidelines-5-Barcodes-Identifiers.
MusicBrainz is certainly more about the music with the credits better detailed and those links to the AcoustID and discID. Whereas Discogs is about selling a unique edition to the collectors. The overlap is really good - and I hope you have this script: GitHub - murdos/musicbrainz-userscripts: Collection of userscripts for MusicBrainz, by various authors Not only is that an importer, but it covers the Discogs pages with pink icons that link back back to MusicBrainz using MusicBrainz own Discogs links.
4 Likes
There is a common reason for the same recording to be slightly different lengths on different releases… If the master is on audio tape, and the tape stretches, or the tape player is running slightly too fast or too slow, then the “Digitally Remastered” CD audio tracks will be slightly different in length.
How many times have they gone back to read the master tapes?
Also, I need to check, but I have two copies of the Information Society’s first album. One pressing has CD+G and the other one doesn’t. But, I recall the packaging was identical, because I took a great deal of precaution to not mix the two up, back when I was ripping my CDs. Anyway, they might have different DiscID’s… or they might be the same. (I’ll look it up later.)
I also have a first pressing of Skinny Puppy’s Last Rights CD, with the 39 second audio mastering error, and a later release with the audio correctly aligned with the TOC track indexes. The first Australian release of the CD had a 4 second offset mastering error too, which was corrected in later releases. The CD packaging is identical for both the error release, and the corrected release.
Oh, I just remembered, I have a CDBaby glass-master replica manufactured CD, with full color silkscreen printing on the disc… with the wrong audio program on the disc itself. It’s a completely different album than what’s printed on the label (and the jewel case). MusicBrainz has both albums, and I was initially confused why the DiscID was matching a completely different album from what I had (printed info).
… So… yeah, CD manufacturing gets screwed up a lot, and sometimes the CDs make it all the way to the customer before the error is caught.
2 Likes
Please do!
I’m really interested in an example for varying TOCs within a release.
But do these two originate from the same release event then? Are the CD media really identical printed? At least the imprints on the matrix will be different, I suppose.
I mean: If anything is re-done, it gets slightly different.
If they run out of covers and print them a new, most probably there are changes in artwork. So this will be a new release, even if they tell nobody and sell it on their platforms with the same release date, isn’t it?
(Yesterday I’ve done so, although all identifiers of the physical medium would have indicated a variant of an existing release.)
But if there are new CDs manufactured - probably slightly different colors and for sure different matrix information - packaged in remaining covers, this would be the same release? Especially if there are audible differences, I disagree.
I said it before: Personally I prefer to have anything with different audible content a different release (and that is even if there are longer pauses).
But I hope you will consider this to be a release by the performing artist, and not by the one on the front cover.
1 Like
I haven’t looked at it it now, but I really will! And, by the way, thanks to all your kind guidance through the jungle of MB. You will be glad to hear, that I’ve decided to reduce the resolution of my uploaded images. 300dpi is really enough to spot the tiniest marks. And even 9600dpi was not sufficient for mould information on some of my CDs. I add microscope images for this information now, which helps to nail down the manufacturing plant. … thanks to your recommendation. 
1 Like
The matrix is different between the two disks. The one with CD+G actually has “CD+G” etched into the matrix around the hub. I think I made scans of it, which I haven’t uploaded yet… if not I get around to scanning it eventually. I’ve setup a light box so I can just photograph the disks, which is much faster and doesn’t have the weird refections the light source in my scanner makes on the reflective surface. (The soft-light looks so much better… Though I’ve also been using a point light source just to make the written and unwritten portions of the disc visible distinct. (Frosted/Matte sandblasted “blank” area, polished mirror “blank” area, and concentric microscopic grooves – the “program” area.))
Anyway, those CDs are in a box in storage right now… so, I’m not sure when I’ll get to them…
Anyway… so after I wrote the above reply while half asleep, I remembered a very good example of two identical releases with different TOCs
I spent an hour or two going through all of the releases on Discogs, trying to figure out exactly what was going on.
This album has a 1:43 long, untitled “Bonus” track… On the earliest US releases Track 17 ends at the end of the song (6:52 in length), and then track 18 is the 1:43 of bonus stuff… On the early UK releases, the bonus track got merged into Track 17, which is now 8:33 in total… But then… later releases of the US release do this too, and leave a 4 second long track 18…
So… the audio program is exactly the same in all cases… but the location of were track 17 ends and track 18 begins has moved around.
And, to make things more confusing… After squinting at all of the photos on Discogs of all the US releases – which match the physical CD and Jewel Case packaging in my hand – the two versions of this CD have the exact same printed materials. (Track 18 being an untitled bonus thing, isn’t actually mentioned anywhere on the covers.)
Both of these TOCs were attached to the same release on MusicBrainz, but I split them into two different releases, because it’s far less confusing this way… (And they could always be merged later.)
I only have this one version of this release, so I haven’t actually checked the minute details from examining the other version. I’m just going off what I can find in Discogs. (And I discovered that someone put the tracks, with the other track lengths, for this album up on Soundcloud, so at least I could check that it was the same audio as the CD I have…)
So… as far as I can tell… all of the printed paper packaging, and silkscreen printing on the CD itself, are identical. The matrix codes on the hub of the disc are distinct (and the disc I have is not in Discogs, but some of the other matrix codes are close).
5 Likes
I have a couple of CDs with new UPC/EAN barcodes, printed on stickers and pasted on top of the originally printed UPC codes. Sometimes the sticker is pasted on the outside of the jewel case, and not actually attached to the paper insert… I’ve been entering these as their own new release, distinct from the earlier one, with the originally printed UPC… There’s no other physical distinction, except possibly the matrix code on the CD itself, which MusicBrainz doesn’t track yet.
So, I can think of a bunch of CDs, which were releases using all kind of TOC tricks for artistic purposes. Mostly hiding a bonus song in the Track 1 pre-gap, or putting 69 one second long tracks of silence to pad out an album to 99 tracks, and put bonus material there. (Or various other long silent pauses.) And then, sometimes, when the album is reissued, or even just distributed in another country. The track list gets messed up and the bonus material is lost, or it’s put in a different location. All of the examples I can think of, do have visible differences on their packaging. Like the differences between the US and UK releases of “Apollo 18” by They Might be Giants. NIN’s Broken EP, um… Tool’s first album… Sarah Mclachlan’s “Fumbling Towards Ecstasy” sometimes has the last track split into two distinct tracks.
In the late 1980’s when CD players were able to do randoms seeks, everyone stopped bothering to support index marks, and so there are a bunch of CDs released back then, for example The song “Domino” on “Invisible Touch” by Genesis, is 10 minutes long with index marks, so you can easily skip to part two… Some later releases of the CD didn’t have index marks.
(It seems that was a feature of CDs mostly only used on classical music releases…)
Since the time displayed when you’re playing a CD is actually encoded into the Q-subcode, you can do all kinds of hilarious tricks displaying any value between 00:00 and 99:99 on the player.
But, anyway… There’s a bunch of albums which have had a “regular” Red Book audio CD release, and then later an “Enhanced CD” with an extra data session… all the examples I can think of have “Enhanced CD” printed somewhere on the packaging though.
I know I’ve read about some CD releases from the 90’s, which were supposed to have a bonus track on them, and people disappointed that there wasn’t a bonus track, because the record company, or pressing plant messed up somewhere… but some discs were manufactured with the bonus material, and some weren’t… and the packaging was exactly the same, and so no one could tell which was which…
I feel like I’ve heard this story before.
3 Likes
I stick to 600dpi as I find 300dpi gets a bit fuzzy on a 50" TV screen.
And always glad to help. It also means I learn more as I go back and read and check details as to what I am saying.
@foxgrrl - liking all the details. I’ve lost count of the number of times I’ve done that “combing through every release on Discogs” madness…
Which reminds me - I have to go and work out how to properly rip this Skunk Ananise release with EAC. They went a little bonkers with the hidden items.
This disc really messed around with the TOC and hid a track BEFORE the first track and other little extras all over the place. I’ll nick some details from Wikipedia:
The album contains a number of hidden tracks and surprises for the listener. The first is hidden before the start of track 1, it is an instrumental mix of the song “100 ways to be a good girl” taken from the first album. After track 3, 7 and 9 there is a short jam (lasting between 0:35 and 1:30; altogether 3:39) which can be accessed directly by rewinding from tracks 4, 8, and 10 respectively. Technically it exists in the pause between the end of one track and the beginning of another. CD players can sometimes be seen counting down to zero while they play these hidden tracks. Some rippers will often append these hidden tracks to the end of the previous song (for example, when ripping the CD).
The final track, “Glorious Pop Song”, ends at 3:43 and is followed by two seconds of silence followed by a recorded conversation which is mostly laughter that lasts for around 30 seconds. Stoosh - Wikipedia
TOC of the extracted CD
Track | Start | Length | Start sector | End sector
---------------------------------------------------------
1 | 2:40.07 | 3:51.50 | 12007 | 29381
2 | 6:31.57 | 3:51.48 | 29382 | 46754
3 | 10:23.30 | 5:04.12 | 46755 | 69566
4 | 15:27.42 | 5:59.60 | 69567 | 96551
5 | 21:27.27 | 3:27.28 | 96552 | 112104
6 | 24:54.55 | 4:13.32 | 112105 | 131111
7 | 29:08.12 | 5:11.13 | 131112 | 154449
8 | 34:19.25 | 4:31.62 | 154450 | 174836
9 | 38:51.12 | 3:14.50 | 174837 | 189436
10 | 42:05.62 | 3:50.28 | 189437 | 206714
11 | 45:56.15 | 4:18.00 | 206715 | 226064
2 Likes
… I know what your talking about. I got a hint that a diffusor on the glass would minimize this effect, but I haven’t found a suitable diffusor. I would need a photographic filter, but they are to small for a CD and diffusors for lighting purposes make too much diffusion - spectral patterns vanish, but so do the details and reflecting silver becomes bright white…
The first one has 3 DiscIDs, the second is currently a separate release - you added it as a separate release because of this differences (and probably because of the Disambiguation comment). Anyway, I support a different release in this case.
The “6:52” release has 3 DiscIDs, but that seems to be not any different from all once merged releases, impossible to be separated entirely. I don’t think they belong to the same release (one of them is a “track 17 - 8:33” disc).
Is it really identical content? Total times are different.
If it would be unchanged content (unchanged artwork, same release event, of course) and there are only altered indices, it would really be a case where I could eventually accept a second DiscID.
But even if it’s identical, it would be hard to have the recordings credited properly. In fact, I would make track 17 of the 8:33 release a “compilation of” track 17 and 18 of the 6:52 release. The 4 seconds silence track 18 would still be it’s own recording.
I wouldn’t do so. Most of the track times are not accurate. Only the TOC of the disc shows the real times.
If it’s from the same manufacturing plant at about the same time, I call it a variant, even if there’s nothing mentioned on Discogs (of course with identical artwork, and that’s sometimes hard to check).
Really? I’ve found not one amongst the CDs I have looked up by now. But I haven’t touched most of my older releases, I would love to have a hidden pregap track.
Ever complicated…
Really? I thought you disliked these huge files. I noticed that downloading all cover art from CAA really takes some time.¹
But I can’t see a gap between 3 and 4, etc. Are these “jams” part of the ripped content? Are they part of one of the tracks? Track length does not indicate this. Where is the data hidden?
¹) There were really no details only showing up on higher resolutions, only scratches and dirt! 
And I thought CD with a diameter of about 1500 pixel would be sufficient even on a 4K display. But if you think 600 is still useful I will continue. Mostly the quality is limited by the overall condition of the the items. It seems to be more useful to edit out ugly stains to have an acceptable Front for upload.
1 Like
I cut a 120mm circular hole in a sheet of acrylic, set that on my flatbed scanner, and then put the CDs in the hole to scan them… Always lined up in the exact same spot… because… I would also make (at least) two scans of the CD, rotated 90 degrees between each scan. So that way I could re-allign the two images, and merge/separate the stripy light reflections of the background, and the printed text and images.
I haven’t actually gotten around to doing the second part yet. I’m thinking of just uploading all of the raw unedited scans (as type “raw”) to the CAA via the appropriately linked releases, and then someone (possibly not me) can do the post-processing.
I wipe off all the dust, and I scan everything at 1200DPI… you can see the CMYK dot screens
… Because I only ever want to have to scan this stuff once. So I’m capturing as much as I can, before I seal all of the original materials up in boxes and put them back into storage.
I also scan the booklets and stuff at 1200DPI four times in four different orientations; Because wrinkles, creases, and folds reflect the light from the scanner in one direction, and are in dark shadows the the opposite direction… so if you scan it one way, flip it around, and scan it the other way. You can align the two images, and the highlights and shadows will exactly cancel out, giving you a completely flat image.
That’s my theory… I haven’t actually done the post-processing on any of this stuff yet either… I was going to automate the process as much as possible, because it’s thousands of images now. I’ve been considering just uploading everything as “Raw”, and someday, someone will process everything into something that looks perfect.
I’m kinda on a slow internet connection at the moment, and the CAA didn’t seem to digest large images very well from when I was uploading some of my early tests.
At 1200DPI you can read the really really small print. Also since the CMYK dot screens are distinctly visible, you can re-separate the colors and process each process color individually. (… in theory.)
Mostly I don’t want to have to get the CDs back out of the boxes for reference, with everything scanned/photographed, anyone can just look at the images.
I did the minimum amount of editing. I think I only moved the single DiscID for the exact disc that I had, and left everything else as it was before.
Yeah, I’ve seen a lot of things in Discogs with very wrong track times.
I only have one singular CD as authoritative ground truth about this release, and needed more clues about why the last two tracks were so different…
I thought there was a report you could run on MusicBrainz to list these, but I guess not. And searching by track number doesn’t seem to work… (though the Regis web search interface. You can do it if you connect directly to Postgres and run a SQL query, but there’s a lot of work involved to do that.)
Here’s a list:
Wikipedia has a couple more lists for CD Bonus tracks in general. (Possibly capture grooves on vinyl releases too.) I’ve seen a few other web sites with lists as well. There’s probably an old Usenet FAQ with most of the same info that everyone’s been copying forever.
5 Likes
That is the problem. I am sure the way I used EAC in 2017 has ripped it in an incomplete manner. I just have not had time to check it. Problem is also exaggerated as my old Marantz CD player refuses to work, so now only have PCs with Pioneer CD players to work with.
Read the Wikipedia quote. That track before track 1 does seem to be implied by track 1 starting at 2:40:07. And the other jams are reached as per the wikipedia notes - you need to rewind when the track clock is at 0:00.
This is also detailed in the MusicBrainz release with the [unknown] bits added to the relevant tracks. Different to other Releases in the same group.
If I can work it out, I’ll post a fresh thread with the differing EAC logs in it. Not had much time today to go into it.
Now is that the one where there are two different grooves in the vinyl so you play a different track depending on where the needle lands?
I remember running a vinyl copy of The Wall backwards to find hidden text.
So many tricks now lost in the Digital World. Important to document as they don’t always make their way onto the re-issues.
600dpi is the suggested guideline (How to Add Cover Art - MusicBrainz). It doesn’t make the file that big to me, but I do upload as an 80% compress JPG as it doesn’t loose too much but keeps quality high.
Take a 80% JPG and a PNG and zoom in on them both. It ain’t too bad as to differences, but JPG will store the data much more efficiently than JPG. (But artwork conversations will have us going OT and the OT Choppers will come and chop this posts)
1 Like
Oh, I thought of another example, the 1997 album “Karma” by Delerium…
Track 10 was the song “Koran” (10:04) on the original release. For all later releases, track 10 has been replaced by the song “Window to Your Soul” (9:25).
For a decade or two, I only had the later release of this CD, with “Window to Your Soul” on it. And then one day in the local record shop. I saw that they had a first edition release with “Koran”. (The tracklist is printed on the back cover.) So, I bought it, excited to finally have this song… and the CD itself had “Window to Your Soul” as track 10, not “Koran”. I was disappointed.
I bought the CD used, so I don’t know if someone swapped a later release of the CD with the first edition release, before it was sold, as used, in the shop… But all of the packaging is identical to the original 1997 release, but the actual program material on the CD itself is the later 1998 release with the other song.
It’s possible the record company changed the master on the second manufacturing run, but didn’t reprint anything else to reflect the change… or someone mixed up the CDs and cases before I bought this used, in a record store.
In theory, the matrix code on the CD hub should provide a clue, right?
Oh… and both sets of DiscIDs were attached to various releases at random.
1 Like
No, thank you! But I will not look at it! If I do, I will instantly know that I have none. It’s not exciting anymore. 
I probably will look up the list, when all my CDs are added in case I overlooked one.
But these are definitely different releases. Matching artwork is only one indicator. If there’s a different release date or place, it’s different regardless of artwork match. And this one has a different tracklist. It hardly can be the same release. As it seems, all later releases had Koran replaced.
It mainly points to the plant, which has manufactured the disc, sometimes to a period of time, but both discs and covers can be reused for a later release. I have a box set, where one disc has the same identifiers like a disc which was part of another box set released six years earlier. (same content on this disc)
But mostly these disc identifiers are really helpful. 
But then there should be something in the data stream at this position. And the hidden content must show up in the EAC ripped file, if you play it as a whole.
I read this once … although I really think 300dpi will hide no details. I do 90% JPG, but 80% would not make it much smaller compared to 25% of the pixels.
On 1200dpi you can see tiny offsets of the print pattern. And every dust particle is enlarged to a monstrous rock. I try to clean every item carefully, but mostly it gets worse.
I don’t miss details on my scans. It’s mostly information on texture what’s lost. Or if it’s thin or thick paper, is it glossy? (what does plasticized look like?)
I’ve tried such things on the above Arcade Fire cover - without any success. My scanner not only causes colorful patterns on reflective areas. Reflections originate of the single LED elements on the scanning bar and are reflected as separate light sources. So there are stripes or a line of circular reflections. Nothing I managed to edit out with scans from different angles. 
EDIT: … I have to reverse my complaint about lost texture. Right now I’ve inspected the scan images of the booklet of Arcade Fire’s Neon Bible box edition provided by @Fabe56 and the difference between my plain paper print and the “plasticized” booklet are clearly to be seen. You would not know, how the physical object looks like, but you clearly see it’s different. 
(I will keep scanning 600dpi resolution, in case… )
I can’t find the CD at the moment. It has been eaten by my house. When I ripped by EAC I had forgotten about these pregap tracks. Once I find it I’ll bring details and will work out how to get EAC to rip it fully.
My current rip starts track one without the pregap hidden track, but you can see this in the track times in my EAC log I posted above. Track one starts at 2:40. Have a look at the other DiscIDs (incorrectly?) attached to that version of Stoosh. Those shorter ones start track one at 2 seconds. No space for the 2:40 hidden track shown in my rip log.
The other hidden tracks\jams are tagged on the end of the previous track as shown in the track listing for that release. (I have just listened back to my FLACs)
This is one of those cases when it would help if I did have multiple copies of the releases. I only have the CD I bought in 1996 and none of the later issues to compare to.
And please stop talking about image editing in this thread. I get in trouble for replying to your off topic conversation. 
1 Like
There are two in the edit history that have been added to all releases (1, 2), you can ask what release they have 
The others, just remove them! Then we have a release that is 100% accurate to what you have in hand and you can quiz the next person to add a discID 
2 Likes
Sorry about that! Didn’t want to.
Some CDs seem to be eaten by my house too. I’m sure, I’ve got them… 
Then the 3 without hidden content can be removed safely as they don’t belong to the original release. Then there are only 2 left.
These two others are quite interesting:
For your release: most probably there was no second master. As I understood, it’s a rare case that the master is replaced during production (except it is faulty). I would think it’s legitimate to remove the other ID. There are still releases left for it…
For the other releases with multiple IDs: it depends on if the credited track times (in this case more the credited tracks) are accurate. If they would be accurate, all DiscIDs not matching could be removed. Possible Problem: if the pregap track was not ripped and thus not credited, it might be a 50:16 DiscID after all.
… this might prove very interesting (and productive) as it was for my Arcade Fire release. We now have two releases with only one discID each. 
…
I wrote the above text before I read your comment where you pointed this out in short
I agree now. I didn’t when the discussion started.
1 Like
@ernstlx digging this discussion back out as I just seen a rather interesting pair of discIDs on a bootleg. Part of the “Can you have two discIDs on the same release?” question.
Have a look at these two:
https://musicbrainz.org/cdtoc/f4JcV9vCOtWDlD6rhW0n89IdpJY-
https://musicbrainz.org/cdtoc/pfyfe5N9e6JADMoX2jhhCg5zsAE-
Open both pages in your web browser. Now flip back and forth. Notice the ONLY difference is the very last digit of the final sector. It is only ONE digit more in length. Not enough to change any real length of the music, but enough for a totally new DiscID.
As this is a bootleg it could be a different copy, or it could just be a rounding error in the reading of the TOC.
The odd thing with this CD is they have crammed as much as they could onto a single CD. A good example though of why you need to check close when discarding alternate DiscIDs.
4 Likes
One is 1/75 of a second short - a separate release!!! 
No, of course it’s not. Thanks for this example!
That’s true, but have given up on it anyway. I only add “Verified DiscID for this release is …” to the Annotation. I thought of cloning a multiple-DiscID release, then removing my DiscID from the original before I add additional information. But I doubt, that all of the remaining even belong to the same release group. And the removing part would cause discussions.
The easiest and quickest way is an annotation and for the database I make sure that track durations for my DiscID are set.
1 Like
Hello again old DiscIDs thread. Today we have a new example for the classroom “Show and Tell”. But this time it is not about Multiple DiscIDs, this time it is about one Manufacturing plant changing owners and keeping the SAME DiscIDs for many years.
I own two near IDENTICAL CDs. Identical boxes, identical booklets, identical painted faces on the fronts of the CDs, IDENTICAL DiscIDs generated by Picard.
BUT they are printed at least six years apart at what is “technically” a different company.
A 1988 MADE IN W.GERMANY BY PDO pressing
And a post 1994 MADE IN GERMANY BY PMDC pressing
It must be post 1994 as it has a SID code. And these both have IDENTICAL DiscIDs generated.
Note that the PDO plant in W.Germany became PMDC in 1992 (and then UNIVERSAL M & L and finally EDC)(*). The same Master has been used for at least six years - maybe more.
In this case I have not made a separate Release. The booklet being printed still had West Germany on it. BOTH CD faces are identical with the same Made in West Germany text printed on the rim. Just the hubs are different. Have a look at the artwork on the above linked release.
In MB eyes the differences seem to more be about visual differences on the booklets and CDs. So I have kept these as a single merged release. Seems to make most sense as only the geekiest of geeks will see (or care about) that little change on the hub.
(Hey - this Geek even notices that both hubs are fully silvered, and the original 1988 release is so old the CD doesn’t even have the ridge moulded onto it!)
( * ) To read up on the histories of the plant in Hanover, Germany one has to go look over at Discogs. They even built a new pressing plant during this time, but took the masters for this CD across to it. I’ll see if I can find a way of re-writing some of those history notes to add them to MB.
6 Likes