I’ve made my first couple submissions and for one of the CDs I’ve added the Disc ID.
The problem is that the entered times from the disc scan is 1 second longer on all tracks compared to online times.
and my scan:
I used cuetools cdripper to do the scan.
Is this a problem with cdripper? It’s the first time I’ve used cdripper. Do I have to do a special setup for the drive before scanning?
EAC shows the track times correctly. Is it possible to add a Disc ID from EAC?
Track times in detail are always a bit tricky. There are several things at play here:
Audio tracks don’t end at exact seconds. If you look at the disc ID entry for track 4 for you see it is 16848 sectors long. At 75 sectors per second that is 224.64 seconds. Now if you only display seconds you have essentially three ways to handle that:
truncate the sub-second part (224)
round mathematically up or down (.5 and above gets rounded up, so 225)
go to the next second (225).
I think I have never seen the third option applied (it’s probably also the least justifiable). But truncation is pretty common, as is rounding. MB and also Picard round mathematically. Both approaches are “correct” in their own way.
You are comparing digital download files vs. times on CD. Those can as well be different.
Displayed track times might be just manually entered metadata. This is for example the reason why track times on the back cover might be different from what your CD player shows.
Not the issue here, but for completeness: Getting exact track times from the digital files is not always that simple. E.g. if you have a raw AAC or AC3 file displayed track times are usually a best guess based on data size.
In this case the one second difference is likely just a different approach to rounding.
If I had set the times, it would mean “no need to check anything” to all the future passers by.
Thus once someone sets real track times, if they notice they don’t match the recordings, they will likely change the recordings at the same time.
It’s an opportunity to fix.
I see. I usually tend to that for physical releases (except vinyls and cassettes). I’ll start leaving them blank from now on so CD owners can set the Disc IDs themselves.
If you can’t get the exact time from the DiscID, I do still think it can be helpful to have the time listed on the medium packaging. Even if the times aren’t perfectly exact, they can be useful in identifying releases or recordings.
Sometimes there are other sources to set the length for a CD. E.g. adding track times from back cover is usually fine. There are cases where they are wrong, but at least it is an official source and if someone has the disc ID it can be corrected later.
I think the benefits of using digital times for a CD release (Assuming: I don’t have access to the CD, and I don’t have any reason to believe that the track times would differ/recordings are different) outweigh the slim chance of a track time being wrong.
If someone comes along with the CD later, and it doesn’t match, they can update it, or create a new recording, no harm done.
So I can see how long a release is, and compare with similar tracks from compilations etc.
If we’re going to assume the times could be really different, why not make different recordings for every release in a group as well, until verified ¯_(ツ)_/¯
I think you already knew, when typing this, that you were exagerating.
Track duration is a minor thing, track entity is more little than recording entity.
There is a big difference between not setting unverified minor data (track duration) and making duplicates of big entities like recordings (with all relationships, ISRC, AcoustID, all that comes with it), it sounds completely overkill.
That’s my point - who cares if a track time is off by 1 second (like in OP’s case). If we care about that we should care about actually important stuff, like potentially using incorrect recordings, more.