I have the cassette version of a release which is listed here in it’s CD release [1], something I digitized almost 12 years ago. One important difference is the cassette release had three extra tracks the CD doesn’t have, so I thought I should fill that out (I have the cassette still, and can compare it to it’s Discogs entry).
I’m wondering how readily the CD info could be copied and then have the three other tracks added (the last track, #19, would become track #22, or ‘side 2 track 11’). Additionally I’m wondering if there’s any value in getting an AcoustID of the tracks, since someone else digitizing their cassette would be using different equipment, different software, and wouldn’t split tracks at he exact same place).
Adding my US version of the Dirty Looks 2-disk set was simpler.
When adding your cassette as a new release, at the second page you are asked if you want to copy another release. You can select the CD track list, and on the tracklist page use the Add Tracks button to add the three extra tracks.
The advantage of this is the recordings are already linked up, the downside is the lengths are all the CD lengths (to 1000ths seconds) and need comparing to your cassette.
On Release Group page hit Add Release in Right Hand Margin…
Update catalogue number, barcode, etc then then hit next and you are on this page:
Which was the other part of my question. Since it’s a tape I digitized myself, the precise length of the recording (especially at 1000th of a second) would vary for someone else digitizing their own copy, and I don’t know if the AcoustID is more like a hash of the file itself (like MD5/SHA256), or somehow analyses the variation of sounds within the file. If the former a scan is effectively useless, if the latter it would actually recognize someone else’s copy.
I would overwrite the lengths on all the tracks. Using your own lengths as reference. You can only type mins:seconds but this is useful as it clears out the fractions that are more true for the CD than your version. Those fractions are generated from the CDs DiscID \ ToC.
With the AcoustID, I would still add them. It is likely you will have slightly different numbers as I am not sure how the silence works at the start of a track. But visually you can see that they are similar if you use the compare.
I haven’t done this with cassettes, but I have digitized LPs, and there is often an acoustid match between my digitized tracks and CD tracks already in the database. So I think submitting acoustids for your cassette would be worthwhile.
Seconding @ivandobsky’s comments on the track lengths. There are cases where cassette and CD releases may use different edits, resulting in significant differences in track lengths.
Unless you think there’s a different edit used on the cassette, I wouldn’t worry much about the lengths. As you have said, they will depend on the playback speed of the deck used, which is far more variable than that of a record player.
Edit: Maybe I should be more clear - If you are tagging your rip against the new MB release you’ve made, using CD times, you will very easily be able to see if something is substantially off. If it comes down to a very slight time difference (e.g. a difference that could be entirely down to where you’ve decided to split the tracks + the speed of your deck) I would leave the CD times, rather than noting the millisecond vagaries of your specific ripping setup.
What I meant was that the millisecond times can be misleading. If you just clone a release at the “Release Duplicates” page you get track times from the CD’s DiscID. Which are misleadingly shown as thousandths of a second(*). Overtyping the times with real times clears this potentially bogus data out and makes it clear the track times are real and have been checked. You can only type in seconds.
(*) Yeah, I realise that you need a userscript to see the fractions of a second. But when trying to untangle multiple releases of a track it is good to be able to rely on the track times. Seeing an LP or Cassette with fractions of a second makes you wonder if the tracks were check at all, or just copied by “Release Duplicate” page.
In general when I duplicate a release and am not sure about durations, I leave the track times empty for this release, to show they were not carefully checked.
When I release a tape on my label I’m using the same master files as for every other medium (even when I do a different master for the medium, the length stays the exact same). I can’t see why anyone would be adding or removing length for cassette tbh.
Unless I’m missing something (possible ) you cannot get more canonical than the CD/digital/master length, length discrepancies from rips are from equipment that is running at the wrong speed.
Edit: As I said earlier though, if there are noticeable discrepancies, it’s time to look closer. I’m not saying it’s always the case!
…are you saying they shaved milliseconds off cassette and LP tracks?
Like I said in both my posts, if there’s a notable length discrepancy, definitely change the times.
p.s. when were cassettes limited to 30 + 30? Genuine question, I’m not familiar with the history, I just assumed pro-duplicated tape was always snipped to length. I don’t cut my tapes to exact lengths because I’m on a budget, I just order the closest general length I need, generally in 10+10, 15+15, 30+30 or 45+45 lengths. I have put different edits on tape before to account for length, but I had to cut off substantial amounts - e.g. a ten minute track turns into a 7 minute track.
My point re. ‘why anyone would be adding or removing length for cassette’ was from a tape producers/manufacturing perspective. Specifically when it comes to milliseconds, or a second or two, in time difference.
I just can’t imagine a scenario where they would use a different length master. Even when I have someone adjust the master for a specific medium (e.g. I don’t want the needle jumping on vinyl, or making it ‘warmer’ for cassette), it doesn’t change the recording/track length.
My personal experience also isn’t universal, I am happy to be corrected
It happens sometimes. CD can be different to LP and Cassette. It is rare, but better to document what you have in hand than copy from a different medium.
I could go digging for times when an LP is different to a CD (Dire Straits Brothers in Arms, Pink Floyd’s Delicate Sound of Thunder) but I don’t have the time to find this right now.
Always best to document what you have in hand than just copy another release.
Yes, obviously… as I said previously I myself have released tapes that have different times to other formats. But never just a couple of seconds, and definitely not milliseconds.
The argument you’ve made is that a label would use masters that are milliseconds different, when the recording and everything else is the same. You’re advocating for documenting speed differences introduced by playback equipment. Any evidence or examples to back this up as a more “accurate” method would be welcome.
As said previously, if the times are substantially different, don’t clone the track times from digital media (e.g. CD).
You have jumped to a lot of conclusions in that post I never suggested. All I was suggesting is to document what you have in hand.
A real world example here. A CD release and Digital Media release use a different version of track 5. Adding the cassette it would be natural to just clone one of these versions, but here it would be important to check lengths to know which track 5 you actually have.
This thread is doing my head in… I’ve specified five times/in five posts that anything other than very short time discrepancies should indeed be looked at very closely.
Earlier in this thread it was claimed that there might be very slight discrepancies in tape master lengths compared to digital formats, down to milliseconds, and I’m saying, from a manufacturers perspective, that I don’t see how that would make sense. If someone wants to shed light on that then please do.
) you cannot get more canonical than the CD/digital/master length, length discrepancies from rips are from equipment that is running at the wrong speed.