ehw
commented
1 year ago Wanted to leave a comment on this and give my thoughts since I deal with this too, since we work with a lot of media on recordable discs.
I think the issue is that regardless whether or not the errors exist consistently, they're still C2 errors which indicate a physical media problem. There will be scenarios where a recorder (or maybe a fault with the media itself) will unintentionally leave mastering mistakes on the physical media itself. From my experience, not every drive model will report the same C2 error, some can read a bit more of the damaged sector than others even if none can totally read the sector fully. So if one drive can read one of the last two sectors differently, the question comes down to - which dump do you choose if you have multiple dumps created from different drives?
As far as redump.org submissions go, I think it will ultimately always be a manual decision that must be made no matter the solution. But, here are some things to consider:
1.) Do the last two sectors exist within the discs TOC? If they don't, then they don't affect any data that's part of any area meant to be accessed under normal use, theoretically. Therefore it doesn't really matter if errors exist or not. 2.) Do the last two sectors affect a data track or an audio track? Data tracks, depending if ECC/EDC is utilized, can be verified and even corrected. For audio tracks, this is trickier as C2 status bits are your only indicator of an issue that's exposed to the user (software). 3.) Can the last two sectors be refined with multiple drives? This is something I'm going to experiment with soon. As I said before, since every drive can read/interpret these kind of errors slightly differently, it might be possible to create a refined dump that best reflects what the original media actually has for those two sectors by combining the results of multiple dumps with various drives. 4.) If the sectors belong to a data track with EDC/ECC data, does the checksum reported in the header suggest a sector with a particular data pattern that can be assumed (like a sector that is all 00'd)? If the C2 status bits returned by the drive confirm that no bits affect the region that contains the checksums, then you could use this to at least determine what the data might possibly be in certain scenarios. However, this might go into "restoration" territory, and not preservation, as such data can always be manipulated or not indicative of the actual media itself.
Just something to think about as maybe this a problem that can be solved.
Tatsh
7Seventy7
Whovian9369
DopefishJustin
superg
Many CD-Rs (betas, press asset discs etc) consistently have Errors in the Last 2 Sectors.
To be specific, these errors are NOT caused by scratches and when dumping with discimagecreator without the /c2 option, you consistently get the same hashes.
The current method to dat such discs in redump with dic is to dump twice and compare hashes. Here's an article about it on the Redump wiki: http://wiki.redump.org/index.php?title=Dumping_CD-Rs_with_Errors_in_the_Last_2_Sectors
Could you do something like this? If CD-R & Errors in Last 2 Sectors, Then (do whatever should be done).