DiscImageCreator

[sarami]

DIC crashes when there are too many C2 errors.

fixed (probably)
http://www.mediafire.com/download/u0olo2c51f2qv13/

[pablogm123]

Unfortunately, it still crashes when ripping a CD with an insane ammount of E32/C2 errors.

And fortunately, DIC can rip from USB drives (actually ATAPI drives converted to USB 2.0 via an ATA/SATA to 2.0 adapter) once and for all.

[sarami]

>pablogm123, Nexy
How much error is there on the disc?
At present, it can fix the "4096" C2 error.
I can increase this values if I can allocate memory.

http://www.mediafire.com/download/u0olo2c51f2qv13/
it can fix 65535 C2 error if it alloc mem.

[F1ReB4LL]

And again, reported all the errors were fixed, but cdmage shows 2 errors in the data track. There's a bug in c2 reporting/fixing somewhere.

[pablogm123]

This is the CD I'm using for testing as extreme sample of damaged CD, an used music CD I got whose metallic layer is awfully damaged by mistreat and moisture.

81705 CU/E32 errors. For PlexTools C2 means E22 error, data damaged yet still recoverable.

While memory usage is high (1312900 KB of private bytes in my system, but that's fine with me because my system has many idle memory), the new release doesn't crash finally when ripping a CD with many errors.

Still testing the re-reading method...

Update, ripping the heavily damaged audio CD:

Command line: discimagecreator.exe -rall g: 8 dump c2 20 fua. Yes, only 20 in order to not to stress heavily my PX-755SA because the tested disc has an insane ammount of damaged zones. As expected, it's impossible (for this disc) to get error-free re-reads (no matter how much you reread the sectors) for every sector and retries are exhausted.

This time has worked properly for me: a secure ripper must detect possible errors (by C2 pointers, for example) and, then, re-read many times the suspicious zones in order to try to reconstruct the good data. If that isn't possible, must, at least, report very clearly that the dump contains still suspicious sectors. Rather than «Reread reached Max   20» I would prefer a clearer text, for example:

Reread reached Max   20
There are unrecoverable errors

Code: Select all

Creating img(LBA) 166954/167523 Detected C2 error. LBA 166956
Creating img(LBA) 166955/167523 Detected C2 error. LBA 166957
Creating img(LBA) 166987/167523 Detected C2 error. LBA 166989
Creating img(LBA) 167003/167523 Detected C2 error. LBA 167005
Creating img(LBA) 167004/167523 Detected C2 error. LBA 167006
Creating img(LBA) 167312/167523 Detected C2 error. LBA 167314
Creating img(LBA) 167426/167523 Detected C2 error. LBA 167428
Creating img(LBA) 167523/167523
Reread times    1, ErrSectorNum 7124/7124
Reread times    2, ErrSectorNum 7124/7124
Reread times    3, ErrSectorNum 6093/6093
Reread times    4, ErrSectorNum 5911/5911
Reread times    5, ErrSectorNum 5733/5733
Reread times    6, ErrSectorNum 5637/5637
Reread times    7, ErrSectorNum 5553/5553
Reread times    8, ErrSectorNum 5498/5498
Reread times    9, ErrSectorNum 5441/5441
Reread times   10, ErrSectorNum 5407/5407
Reread times   11, ErrSectorNum 5372/5372
Reread times   12, ErrSectorNum 5336/5336
Reread times   13, ErrSectorNum 5296/5296
Reread times   14, ErrSectorNum 5271/5271
Reread times   15, ErrSectorNum 5242/5242
Reread times   16, ErrSectorNum 5218/5218
Reread times   17, ErrSectorNum 5197/5197
Reread times   18, ErrSectorNum 5167/5167
Reread times   19, ErrSectorNum 5156/5156
Reread reached Max   20
Move .scm to .img
Creating bin, cue, ccd(Track)  7/ 7
Calculating hash. D:\\Redump\DIC\\dump (Track 7).bin
End -> 2013-11-08(Fri) 20:10:08

Logs of this ripping process: https://www.dropbox.com/s/t3oqxp2adhkhvuo/Log.7z

And USB drives (mines are actually ATAPI attached to a ATA/SATA to USB adapter) work properly newly.

[Nexy]

F1ReB4LL what disc is that? You know there can be no C2 error but still be intentional header/EDC/ECC errors right? (some psx disc for example with last sector of data track)...

Anyways, better make it so DIC can handle absolute worst case scenario for errors, I guess with some GC disc this is gonna be like 450000-500000 sectors? Maybe you should go back to using files for C2 to avoid excessive ram allocation, or make it sliding window.

Disc I have was having > 80000 error (worst case test).

Someone reported that multiple data tracks weren't dumping proper... seems to work fine for me (D!Zone 3).

Are there hard coded fixes for certain discs in the code? Somebody said there was. I hope you didn't do that, if there is please remove such things and implement proper fixes for them. We do have a couple of other coders you could talk to.

[sarami]

fix C2 option
http://www.mediafire.com/download/u0olo2c51f2qv13/

Code: Select all

                c2: Fix C2 error existing sector (Only CD)
                        val1: value to reread (default: 1024)
                        val2: value to fix a C2 error (default: 65535)

And USB drives (mines are actually ATAPI attached to a ATA/SATA to USB adapter) work properly

umm.. I don't fix about this.

reported all the errors were fixed, but cdmage shows 2 errors in the data track. There's a bug in c2 reporting/fixing somewhere.

Depending on a drive, it's possible that it don't report C2 error exactly.
Therefore, in this tool, multiple read a non C2 error sector (at present 2 times)
If necessary, I can change a reading times.

[pablogm123]

More suggestions:

Get exclusive access to the drive when ripping (like many burning programs do when burning), so that another programs cannot disturb the drive and the re-reads.

---

The possibility to set different read speeds for the initial read and the readings. For example, initial read at 24x speed and re-reads at 4x speed.

---

Rewrite the -ra module so this tool can act as a CDTool replacement to dump as audio (D8 or audio trap disc) any range that the user desires with subcode and C2 error reporting (outputting both as .sub and .c2e files), adding the possibility to correct the offset (with the possibility to override the standard one). This module should support Plextor and non-Plextor drives and could be useful to dump the high density of DC games.

If a Plextor drive is used, D8 command and order is Main+C2+Sub.

If a non-Plextor drive is used, BE command and autodetect if the drive uses Main+C2+Sub (NEC based for example) or Main+Sub+C2 (Mediatek bases for example).

For example, to dump this disc https://redump.info/disc/5626/ using my PX-W4824TU after hot swapping, something like this one:

discimagecreator.exe -ra h: 8 add +13 45000 549150

[F1ReB4LL]

Depending on a drive, it's possible that it don't report C2 error exactly.

Mine is PX-760.

[pablogm123]

I strongly doubt that PX-755/760 drives sport inaccurate C2 error reporting:

The UltraPlex PX-40TS returns reliable C2 information but can only report when a C2 error occured in a frame, not where. Reliable C2 information is a must. Without it, PlexTools cannot approach EAC's accuracy. The recent Plextor drives starting from model 12/10/32 have balls on accurate C2 information and form a very powerful pair combined with PlexTools. Plextor recoders (not cd-roms!) before model 12/10/32 have rather inaccurate C2 information.

Code: Select all

http://club.myce.com/f43/post-your-experiences-latest-plextools-1-15-a-53228/#post308756

In fact, it's possible that these drives prefer stay on the safe side and mark good bytes as bad bytes.

It's a natural property of RS codes that the more bytes you try to correct,
the more likely you can get miscorrections, so the (x,4) strategies should get
lower scores at C2 accuracy tests. However, these tests should also take
into account bytes that are correct but are flagged as wrong, as there are
also several strategies there, and of course the more bytes you flag as
wrong, the easier you can reach a high score.

Code: Select all

http://club.myce.com/f61/c2-accuracy-vs-efficiency-they-exclusive-67558/#post411413