1 (edited by topkat 2007-08-12 20:47:16)

Hello everybody!

I'm new to this project and want to contribute soon.

I have some PAL-Games lying around (original, of course) that arn't in the db yet.

Before I start dumping I have a few question:

1) What exactly is EDC? I guess it's something like "Error Detection Code'? As far as I know, the PSX doesn't use bad sectors for copy protection, what's the deal with EDC and PSX-Games?

2) There are some disc with a data track (track 1) and an audio-track (track 2) with an length of 03:32:00. The crc of the audio-track is 7976083e. The same track occurs among different games, e.g.

- Breath of Fire IV (E) [SLES-03552]
- Music - Music Creation for the PlayStation (E) [SLES-01356]
- Dino Crisis (E) [SLES-02207]

and some others. The track actually has no audio in it, it seems to be completly zero (silence). What's the propuse of this track?

3) Why not simply use some kind of image-format (like alcohol or clonecd or the like) instead of the tricky and time-consuming dumping-process? I dumped a game already in the db, and got matching hashes. Then I created an image using the lates DT of the same game. Afer that, I compared both, the dump and the image, and they seem to be identical. I you are interesset, I can post some details.

4) Before dumping I check discs with VSO Inspector, an freeware disc-testing app. On every discs, there are some errors in the first sectors (0 - 16). Is this the area where the copy protection is stored? Nevertheless I got perfect dumps with the same hashed as stated in the db, so i guess that's normal?!

Hope these question aren't too stupid smile

Hi and welcome to our project. I'll try to answer to your questions.

topkat wrote:

1) What exactly is EDC? I guess it's something like "Error Detection Code'? As far as I know, the PSX doesn't use bad sectors for copy protection, what's the deal with EDC and PSX-Games?

Yes, you're right, EDC is error detection code located in each sector, but some old games are missing it from Form 2 sectors, therefore it's not impossible to fully check them for errors. Moreover, if dump which is missing EDC in Form 2 sectors is written to the disc not in RAW mode, EDC will reappear, so dump made from such copies will have different checksums. These images are spread over the internet, so always check EDC status before asking for a patch wink

topkat wrote:

2) There are some disc with a data track (track 1) and an audio-track (track 2) with an length of 03:32:00. The crc of the audio-track is 7976083e. The same track occurs among different games, e.g.

- Breath of Fire IV (E) [SLES-03552]
- Music - Music Creation for the PlayStation (E) [SLES-01356]
- Dino Crisis (E) [SLES-02207]

and some others. The track actually has no audio in it, it seems to be completly zero (silence). What's the propuse of this track?

It's dummy audio track, used in some Capcom games probably to avoid reading errors in the end of the disc.

topkat wrote:

3) Why not simply use some kind of image-format (like alcohol or clonecd or the like) instead of the tricky and time-consuming dumping-process? I dumped a game already in the db, and got matching hashes. Then I created an image using the lates DT of the same game. Afer that, I compared both, the dump and the image, and they seem to be identical. I you are interesset, I can post some details.

Try dumping one and the same disc with audio tracks on two different drives in CCD, then compare checksums - you'll be surprised wink For more info read here and in our dumping guide.

topkat wrote:

4) Before dumping I check discs with VSO Inspector, an freeware disc-testing app. On every discs, there are some errors in the first sectors (0 - 16). Is this the area where the copy protection is stored? Nevertheless I got perfect dumps with the same hashed as stated in the db, so i guess that's normal?!

Most probably this tool doesn't works with Mode 2 Form 2 sectors and reads data in cooked mode - sectors 12-15 are Form 2, and they can be read in RAW mode only.

I hope you're satisfied with my answers smile

3 (edited by topkat 2007-08-14 16:16:50)

Thanks a lot for the infos!

Dremora wrote:

Yes, you're right, EDC is error detection code located in each sector, but some old games are missing it from Form 2 sectors, therefore it's not impossible to fully check them for errors. Moreover, if dump which is missing EDC in Form 2 sectors is written to the disc not in RAW mode, EDC will reappear, so dump made from such copies will have different checksums. These images are spread over the internet, so always check EDC status before asking for a patch wink

I guess you mean 'it's not possible to fully check them for errors' ?!

Is this the issue of the old mastering equicment from Sony producing EDC with wrong reported values? Somewhere I read about this topic and years ago everbody though that this was the copy protection, but as far as I understand this was just a manufacturing failure.

Dremora wrote:

Try dumping one and the same disc with audio tracks on two different drives in CCD, then compare checksums - you'll be surprised  For more info read here and in our dumping guide.

Your are talking about that offset-thing, right?!

OK, here is what I did:

As mentioned, I followed the dumping guide and dumped a game (with audio track) of my collcetion that was already in the DB  The hashes matched. After that i created an image using the latest Alcohol 120 %. Then I mounted both, the dump (via the cue-file) and the mds-image from Alcohol to a virtual drive. To compare both, I used a programm named CD-R Verifier. Here's the resault

- The Mounted dump and the mounted image seems to be identical
- Neither of the both match the orignal disc
- Burned back to a CD-R, it neither match the dump, nor the image and of cause not the orginal (maybe a burner and/or software issue)

As far as I know, the programm reads every bit of a given disc and calculates a checksum. So, if the checksums of the two discs match, the should be (bit)-identical.

Due to low on storage space I deleted the whole test files, but now I got a new hd and will repeat the tests again. Furthermore I will test if it is possible to get a proper dump for the mounted image along with some other tests...

BTW: Is there some kind of recommondation for a good cd-drive? ATM i'm using a LG GSA-4166B along with a small army of different other cd/dvd-readers and writers. Some older drives seems to have better have error corrections capatibilies than the LG. Is that Plextor Premium everbody used to talk about, really that good? It's still cost around 100 €, that's a little fortune for a 5 year old drive...

As mentioned, I followed the dumping guide and dumped a game (with audio track) of my collcetion that was already in the DB  The hashes matched. After that i created an image using the latest Alcohol 120 %. Then I mounted both, the dump (via the cue-file) and the mds-image from Alcohol to a virtual drive. To compare both, I used a programm named CD-R Verifier. Here's the resault

- The Mounted dump and the mounted image seems to be identical
- Neither of the both match the orignal disc
- Burned back to a CD-R, it neither match the dump, nor the image and of cause not the orginal (maybe a burner and/or software issue)

As far as I know, the programm reads every bit of a given disc and calculates a checksum. So, if the checksums of the two discs match, the should be (bit)-identical.

BTW: Is there some kind of recommondation for a good cd-drive? ATM i'm using a LG GSA-4166B along with a small army of different other cd/dvd-readers and writers. Some older drives seems to have better have error corrections capatibilies than the LG. Is that Plextor Premium everbody used to talk about, really that good? It's still cost around 100 €, that's a little fortune for a 5 year old drive...

Checked game probably has dummy audio track. Therefore hashes are identical
And any drive has write offset,  Therefore after record you have received other result.

http://forum.psxdb.com/viewtopic.php?id=641 - about CD/DVD drives

topkat wrote:
Dremora wrote:

Yes, you're right, EDC is error detection code located in each sector, but some old games are missing it from Form 2 sectors, therefore it's not impossible to fully check them for errors. Moreover, if dump which is missing EDC in Form 2 sectors is written to the disc not in RAW mode, EDC will reappear, so dump made from such copies will have different checksums. These images are spread over the internet, so always check EDC status before asking for a patch wink

I guess you mean 'it's not possible to fully check them for errors' ?!

Yes, that's exactly what I meant smile

topkat wrote:

Is this the issue of the old mastering equicment from Sony producing EDC with wrong reported values? Somewhere I read about this topic and years ago everbody though that this was the copy protection, but as far as I understand this was just a manufacturing failure.

Yes, more recent PSX discs always have EDC. The absence of EDC was just a manufacturing error.

About Alcohol: I don't want to use any proprietary formats; moreover, there is no sense in calculating checksums of such image (with data, subchannels and additional disc info stored in one file).

Ok, please get me right, I think this is a great project, esp. regarding the masses of rips (mp3, re-encoded waves and such crap) floating around...

I just though as theses images-formats are able to handle the latest copy protection, it should be possible to get exact dumps. But probably I'm complete wrong wink

I will now testrun your guide with some of my discs and when I 'mastered' the techniqes, I will send in some additions to the db!

One more question: Do the so called 'Platinum Editions' have the same Game-ID as the original releases? Are there any known differences between an original release and the later pubilshed 'Platinum Editon'? (aside from different artwork)

Eg. Resident Evil 2 (G) (Disc 1) (Leon Disc) [SLES-00974] is listed as Platinum, but I own the original release with the same SLES-No. Still I was able to get a dump matching the hashes of the listed Platinum.

ID number identical, but versions can differ.
RE 2 Platinum and Original are probably identical. For this reason we also bring the version in the comment.

Most platinum versions do not differ, some differ significantly. If possible, always note which edition of the game you have.

Was the game you tested your alcohol test with a game with just one audio track ? Could it be that that track contained only zeroes ? In that case drive offset will not really come into play in most drives (some may have some garbage in the first few samples though, but that's not the norm).
While Alcohol and some others can "handle" most copy protections, never equate that with making an exact dump of the data. Copy protections would usually not be dependant on a particular drive offset (since the protection can't really know what drive it is currently running on), so that's irrelevant for Alcohol to get right (... for now; theoretically one could add a calibration track to the disc, calculate the drive's offset from that, and then use that to check against some custom write offset for copy protection purposes, but AFAIK nobody has done that (yet).

Is this the issue of the old mastering equicment from Sony producing EDC with wrong reported values?

Well, technically it produces no EDC at all, i.e. the EDC field is left blank (or you might look at it as an EDC with 4 zero-bytes as its content, which is highly improbable).

Before dumping I check discs with VSO Inspector, an freeware disc-testing app. On every discs, there are some errors in the first sectors (0 - 16). Is this the area where the copy protection is stored? Nevertheless I got perfect dumps with the same hashed as stated in the db, so i guess that's normal?!

That is indeed the case. The EDC in those XA form 2 sectors is zeroized, which does not match the data at all; any self-respecting checksum-checker should complain about that, and that's probably what your tool is doing. It might be interesting to try that tool on a disc without EDC, or rather, the broken EDC mentioned above wink It should really only be sectors 12-15 though (sectors 162-165, if you count the lead-in).

The dumping process is indeed a little convoluted, but in so doing it tries to preserve all the elements of the original disc as best as possible (i.e. not just taking drive-read-offset into account, but also write-offset from the manufacturer. The latter can be a royal PITA to detect properly on many discs that sport large negative offsets, however (since you'd need a drive with a large positive read offset to even see the mentioned garbage data).

norman wrote:

It should really only be sectors 12-15 though (sectors 162-165, if you count the lead-in).

Pregap wink Lead-in is usually 4500 sectors.

10 (edited by topkat 2007-08-26 00:22:59)

Sorry for the late replay. had some problem with my pc (IE constantly crashes when opening the forum sad ) Now using Firefox!

@ norman

Yupp your're right, it was a game with the 'Capcom Dummy Track' (silence), so I guess it would be different with normal tracks. But let's forget about this image vs dump stuff and do some work..

I made some new dumps of my PAL games:

Cybertiger [SLES-02370]
Army Man Operation Meltdown [SLES-02855]
Simpson Wrestling [SLES-03401]

What to do next?