Your CD disc is possibly a different pressing to the one(s) stored in AccurateRip. (Can - Soon over Babaluma - "spoon CD 010" bought in Germany)

I've seen many CDs of mine were some tracks (not the first/last tracks) were different from the ones in the database. Yet, I'm totally sure that my setup and my CDs (happened on tracks with "Track quality 100.0 %") are ok.

What's wrong? How does AccurateRip identify the CDs anyway? By TOC? In case it's TOC only then I think that's not enough. For example what happens with the above mentioned CD, will it get a seperate entry and people who also have that issue will be presented my checksums?

There's often different pressings of the same CDs. You just happened to get a different pressing than the one in the AccurateRip database.

Yes, exactly. I'm sure this is the case with my CDs... but IMHO that's the problem. Apparently AccurateRip can't distinguish the different pressings because it identifies the CDs by the TOC/track lengths/gaps/whatever. It seems it's not enough for some albums.

It has to, else you can't get the proper offsets for your drive. Different pressings can have different offsets, so your drive's offset must only be calculated with the same pressing. It may not necessarily be the case for every CD, but it often is. Better safe than sorry.

Maybe your setup and CDs are in fact OK, but "Track quality 100.0 %" is not conclusive evidence.

AccurateRip identifies the different CDs by looking at their TOCs in the lead-in Q subchannel. This system is much better than e.g. freedb's discid's, which just is a number calculated by the CDs track number count and lengths, and where collisions are frequently occuring. Different pressings of the same CD will have identical TOCs, but will just sometimes be shifted by a number of samples against eachother. This also means that AccurateRip can not know if there suddently is a new pressing of a CD introduced into the system, or just a bad rip of the systems present pressing, but as AccurateRip keeps the different pressings submitted side by side, then if someone later comes along and rips the exact same pressing as yours, then that pressing will now be perfectly identifiable to others...

AccurateRip can't handle different pressings of the same album. It only works if someone else have already ripped the same pressing.
It's logical. A different pressing is not a bit perfect copy. Thus AccurateRip says that the rip is inaccurate compared to the one in its database.

Ah, but in case someone else with the exact same pressing as I have, does AccurateRip then uses my results for comparing them to his or does it use all the other results (including mine of couese), too? In other words: are there many, seperate database "entries" for a single disc ID (in case a totally different pressing is detected "Your CD disc is possibly a different pressing...") or are all results thrown into the same disc-ID-basket?

Does it support multiple pressings or not? Or is it strictly TOC based and the pressings with the most submissions win and the others are then displayed as being <quote>not accurate</quote>.

For a given TOC, it displays how many users have already got the same result as you.
For example if 8 people got the result A, and two other got the result B, and you have the same pressing as them, it will display "Accurately ripped (confidence 2)".

Different pressings have identical TOCs, track lengths, gaps, just the audio data is different. AccurateRip can handle multiple pressings with the same TOC, ie side by side.