DiscImageCreator

[Jackal]

This disc had the first 2 LibCrypt sectors undetected: https://redump.info/disc/50247/

Logs: https://drive.google.com/open?id=1ClsZ6 … 32TtOFAaxd

Maybe it's better to stop using prefixed ranges, and just not fix any subs in case /ns is used? because it happens a lot that some LibCrypt/SecuROM data goes undetected.

Or set a limit to how much corruption is fixed in each Q-channel sector. Now it fixed 4 bytes total, including 2 in CRC. I doubt that read corruption can occur on so many bytes at once, and also in both CRC bytes?

[sarami]

Greatly changed the range.
libcrypt: 14200 - 16200 or 42000 - 44400 => 10000 - 20000 or 40000 - 50000
SecuROM OLD: 30800 - 34800 or 40000 - 45800 => 30000 - 50000
SeruROM NEW: 5000 - 20000 => 5000 - 25000

it's better to stop using prefixed ranges

It's verbose that all sector check.

just not fix any subs in case /ns is used?

If random error exists on the boundary of track, it can't split the img to bin correctly.

Or set a limit to how much corruption is fixed in each Q-channel sector. Now it fixed 4 bytes total, including 2 in CRC. I doubt that read corruption can occur on so many bytes at once, and also in both CRC bytes?

It may take a lot of time to code it and occur newly bugs. You and I don't want it.

[iR0b0t]

sarami, i came accross a weird behavior with /ns switch on this disc https://redump.info/disc/50400/

It affects subchannels, and ALL pregaps are by 1 frame shorter if /ns switch is NOT being used.

LOGS (available for 2 days)

[sarami]

dic checks the subchannel offset by reading 1st LBA of the data track.
SecuROM 99 error version is judged that the offset is +1. ([00:02:01 - 00:02:00 = 1 frame)

Code: Select all

LBA[000000, 0000000]: P[ff], Q[410101000000000002013813]{ Data,      Copy NG,                  Track[01], Idx[01], RMSF[00:00:00], AMSF[00:02:01]}, RtoW[0, 0, 0, 0]

If dic judges the subchannel offset exists in this drive and you don't use /ns switch, it reads from LBA -1.

Code: Select all

LBA[000000, 0000000]: P[ff], Q[410100371045000002000793]{ Data,      Copy NG,                  Track[01], Idx[00], RMSF[37:10:45], AMSF[00:02:00]}, RtoW[0, 0, 0, 0]

If you use /ns switch and SecuROM 99 error version is detected, dic reads from LBA 0 (-1 + 1 = 0).

Code: Select all

LBA[000000, 0000000]: P[ff], Q[410101000000000002013813]{ Data,      Copy NG,                  Track[01], Idx[01], RMSF[00:00:00], AMSF[00:02:01]}, RtoW[0, 0, 0, 0]

Therefore, if you don't use /ns switch, the boundary of the track is shifted by 1 frame. As a result, the size of track 01 increases 2352 and the pregap of track 02, 03... decreases 1 frame and the size of last track decreases 2352.

[iR0b0t]

Is that correct though?

[reentrant]

Interesting I always thought that /ns option should produce identical dumps...

[sarami]

SecuROM 99 error version can't dump correctly without /ns. Is this not good enough?

Puzzle Bobble 4 (T-42301M) - (I have 2 copies of this game):
The dump seems correct from the "disc 1".

I tried Puzzle Bobble 4.
Unfortunately, track 12 is the different hash every time. (all track except for the track 12 is same hash.)
1. TS-H352C FW:NE02

Code: Select all

crc="a34b4e63" md5="07ac424b0e3b45fdcd0c1c65ded5cc50" sha1="b0cab18080ee2c3232c1bcf8d2ae1dadb55d0226"

Code: Select all

crc="993cd51b" md5="93bc67101efc677aa961bc5c77afb4f1" sha1="46abd792c654b1261cadf54736157c73dbc168a6"

2. TS-H353A FW:BA08

Code: Select all

crc="2e347ba4" md5="e57f6d9ab99690b7b1256cd56b00402e" sha1="0ccc6583a6596a4028dafa692d229c5c5410dc4a"

Code: Select all

crc="07a0f42c" md5="dff224d27e57c0a4829adc3264752bd0" sha1="b5f0b5e1b957250c5010351b0103520a71644559"

I don't have Death Crimson OX, but I think it becomes the error same as you.

I don't know whether the problem of the drive or that of the disc. But SH-D162C, SH-D162D, TS-H352C and TS-H353A are similar drive, so using other drive in this list /viewtopic.php?p=3524#p3524, it may work (but I don't guarantee it. If it not work well, Puzzle Bobble 4 and Death Crimson OX may have mastering error sectors).

[reentrant]

What is SecuROM 99?

[sarami]

enum.h

Code: Select all

    securomV1, // a.k.a SecuROM OLD ((8 shifted RMSF/AMSF + 1 error) * 24 times = 216 sector error)
    securomV2, // a.k.a SecuROM NEW ((8 shifted RMSF/AMSF + 1 error) * 10 times = 90 sector error)
    securomV3, // a.k.a SecuROM NEW ((8 shifted RMSF/AMSF + 1 error) * 11 times = 99 sector error) 
    securomV4, // a.k.a SecuROM NEW (LBA -1 + 10 random error = 11 sector error)

[jhmiller]

in the last 2 Dreamcast dumps that I have tried, the DIC shows this error:
Checking SubQ adr (Track)  1/ 1
Checking SubRtoW (Track)  1/ 1
Reading DirectoryRecord    1/   2End time: 2018-04-07(Sat) 15:16:01

Or this:
Reading DirectoryRecord    1/  19End time: 2018-04-07(Sat) 14:32:35