DiscImageCreator

[F1ReB4LL]

Yes, works good here. But you list a number of sectors with c2 errors ("Reread times  22, ErrSectorNum    7/   7"), while, as I've already said earlier, it's better to show a number of error bytes per sector according to the 294-bytes C2-error part in each sector and to take the good bytes from each rereading, instead of trying to reread the entire sector without errors.

For example, there are 5 C2 errors in the sector #123 (bytes 2, 5, 10, 15 and 20 are corrupted).
1st reread: 4 C2 errors (bytes 5, 10, 15 and 20 are corrupted, byte 2 is good - you replace it in the dump) - 4 errors left.
2nd reread: 8 C2 errors (bytes 2, 3, 4, 5, 6, 15, 20, 25 are corrupted, byte 10 is good - you replace it in the dump) - 3 errors left.
3rd reread: 7 C2 errors (bytes 2, 3, 4, 5, 10, 15 and 20 are corrupted, nothing fixed, nothing to replace) - 3 errors left
4th reread: 10 C2 errors (bytes 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10 are corrupted, bytes 15 and 20 are good - you replace them in the dump) - 1 error left
etc.

[pablogm123]

This is the exact method used by the PlexTools program, the mode 5.

Recover the best bytes (least errors) per sector

When Uncorrectable Errors are found, the damaged area will be read again for a number of times (Retries) at the selected speed. When the errors stay, the bytes with the least errors will be used to re-assemble the original data.

http://wiki.hydrogenaudio.org/index.php … ry_Options

[F1ReB4LL]

This is the exact method used by the PlexTools program, the mode 5.

Btw, sarami could also implement 2 different modes: 1 like now, tries to reread the sector until there's no c2 errors and 2 like I've described and like it's described in this wiki, to take the good bytes from the sectors with c2 errors.

And, please, move the rereading number to the commandline, 750 isn't always enough.

[Nexy]

May have discovered part of others issues.

You should probably add some kind of initialization command before dumping. Sometimes it will just abort with a drive not ready error. I have to open isobuster and hit refresh a couple of times before it works again.

The hashing takes ages too, unoptimized routines there? Will you improve them?

[pablogm123]

@Sarami: I would like that you could implement an option to skip the splitting of the dump and the unscrambling of the data tracks.

So that this tool could act as a cdtoimg-d8 replacement, but with automatic offset correction (or manual when you implement the option to override the automatic offset correction for dumping certain audio discs), subcode dumping and C2 error reporting (and rereads with C2 pointers taking into account).

And to defeat the drive's cache when performing rereads, just use the special FUA command as suggested before. I think that this is the best option to flush the cache if you could implement it.

This post:

/viewtopic.php?p=23374#p23374

And this tool to test if your drive supports FUA (you have to insert an audio disc in good condition):

Code: Select all

http://download.cdfreaks.com/download/155

If supported you will see:

[+] Plextor flush command: accepted
[+] Plextor flush tests: 20/20

P.S.: I agree, the hashing is somewhat slow compared with another utilities.

[sarami]

>F1ReB4LL
I examine that fixes per a byte.
By the way, is there an offset for c2 error byte?
If an offset exists, is it different every drive?

>Nexy
>>Sometimes it will just abort with a drive not ready error.
If possible, please tell me the procedure to let it reappear in detail.
>>The hashing takes ages too, unoptimized routines there? Will you improve them
Please tell me the good open-src library.

>pablogm123
>> override the automatic offset correction
It will add at next wip.
>> drive's cache
Tested my drive. My 755sa don't support fua on 0xd8 cmd. I don't know why that is.
Therefore, even if I coded it, I can't test it that its code is correct.

Code: Select all

C:\cachex.exe -i m:

CacheExplorer 0.8 - spath@cdfreaks.com

Drive on M is  PLEXTOR  DVDR   PX-755A   1.08

[+] Buffer size: 2048 kB, read cache is enabled
[+] Supported read commands: BEh 28h(FUA) D8h

Code: Select all

C:\cachex.exe -p -c -r 0xd8 -n 20 m:

CacheExplorer 0.8 - spath@cdfreaks.com

Drive on M is  PLEXTOR  DVDR   PX-755A   1.08

[+] Plextor flush command: rejected
[+] Testing cache line size:
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors

[pablogm123]

Certain drivers and filter drivers, like the sptd.sys, can break that. Due to that and to avoid any possible kind of conflicts (like this one officially reported by Microsoft), in my system I don't install any software which installs upperfilters, lowerfilters and virtual drives, only software which uses the standard SPTI layer, nothing more.

Therefore, the FUA support (if ever implemented) to defeat the cache should be optional, via an additional command line parameter.

My other drive, even connected via an USB 2.0 enclosure supports it:

Code: Select all

C:\>C:\cachex.exe -p -c -r 0xd8 -n 20 h:

CacheExplorer 0.8 - spath@cdfreaks.com

Drive on H is  PLEXTOR  CD-R   PX-W4824A 1.07

[+] Plextor flush command: accepted
[+] Plextor flush tests: 20/20
[+] Testing cache line size:
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors
 1169 kB / 509 sectors

By the way, did you run this program from a elevated cmd.exe process?

[DJoneK]

Opened command prompt as admin

Commandline: cachex.exe -p -c -r 0xd8 -n 20 p:

Code: Select all

    
    Drive on P is  PLEXTOR  CD-R   PX-W5224A 1.04
     
    [+] Plextor flush command: accepted
    [+] Plextor flush tests: 20/20
    [+] Testing cache line size:
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors
     581 kB / 253 sectors

Windows 7 64bit SP1
Daemon Tools Lite v4.46.1.0328
SPTD v1.83

[Nexy]

Sarami, you need to buy some SecuROM protected discs. Dumps are completely inconsistent... BUT this is not an entirely bad thing... apparently you stumbled on a way to dump the twin sectors (this is a good thing) if it can be made consistent it will be great. Both me and MrTikki have run across this now and we discussed it on IRC for awhile today.

On the bad side, all SecuROM dumps are bad an need redone.

F1ReB4LL's idea is to read the disc linear in both directions , 0-end and end-0 to try and capture all twin sectors. But for some reason DIC reads the twin sectors at random.

Also, can you add a switch to save the unsplit scrambled dump from being deleted.

Also, error is "Unable to Access Device" happens randomly, I don't know what causes it. Other tools don't have this problem.

cachex mem test for fun.

Code: Select all

Installed Software

Alcohol 120% 2.0.2 Build 3931
CloneCD v5.3.1.4
DVD Decrypter
EAC .99 PreBeta 5
ISOBuster v3.1
Nero Burning Rom v10 (nerocheck.exe removed)
Nero DiscSpeed 11
OptiDriveControl 1.51
UltraISO Premium v9.51

SPTD Layer v1.81

cachex.exe -i -p -c -r 0xd8 -n 20 d:

CacheExplorer 0.8 - spath@cdfreaks.com

Drive on D is  PLEXTOR  CD-R   PX-W5224A 1.04

[+] Buffer size: 2048 kB, read cache is enabled
[+] Supported read commands: BEh D8h
[+] Plextor flush command: accepted
[+] Plextor flush tests: 20/20
[+] Testing cache line size:
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors

cachex.exe -i -p -c -r 0xd8 -n 20 e:

CacheExplorer 0.8 - spath@cdfreaks.com

Drive on E is  PLEXTOR  DVDR   PX-760A   1.07

[+] Buffer size: 2048 kB, read cache is enabled
[+] Supported read commands: BEh D8h
[+] Plextor flush command: accepted
[+] Plextor flush tests: 20/20
[+] Testing cache line size:
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors
 581 kB / 253 sectors

[pablogm123]

Good news! SPTD no longer breaks the FUA functionality, as was reported a time ago.

And I agree with the idea of keeping a copy of the scrambled dump, without splitting/unscrambling anything. For example, instead of deleting the original dump renaming it to .scrambled would be a good solution.

================

Regarding SafeDisc there is additionaly a little big problem. Different drives based on different chipsets will give you non-constant results, due to the different error treatment implemented by each different brand of chipsets. One given drive can extract uniformely one SafeDisc disc into scrambled form, but the bad sectors will differ from the ones dumped by another drive.

As sample: the first 15000 sectors of a SafeDisc 2 extracted using CDTool and audio trap disc. While each drive can obtain uniform scrambled reads, they differ from the another drive. Only filling these sectors with an arbitrary pattern (like 0x55) can deliver uniform dumps, but unfortunately this is neither a perfect nor elegant solution.

Code: Select all

https://www.dropbox.com/s/rmcnp1l76io1uzc/NEC.7z
https://www.dropbox.com/s/uzkemlu2ks01s6v/Sanyo.7z
https://www.dropbox.com/s/z4v4cj5bwsw4roh/Sanyo-processed.bin
https://www.dropbox.com/s/llwr0wzw5ywudja/NEC-processed.bin

NEC: Pioneer DVR-107D.
Sanyo: Plextor PX-W4824TA.