• Registered: 2012-09-18
  • Posts: 22

Hi everyone, i was wondering if there's possibility to use 2448 standard for PlayStation images. You well know that some ps1 europe games have libcrypt protection and a simple cue/bin in mode2/2352 couldn't handle the protection correctly. In recent homebrew scene of ps3 has been developed a payload by aldostools:

https://github.com/aldostools/COBRA/tree/master/485

It supports ps1 different image formats:

[ Aldo's changes ]
Improved PSX media type support, now it supports following disc types
2048 // 2352 // 2336 // 2448 // 2328 // 2340 // 2368 (PLAYSTATION and regular CD ISOs)

According to what seen on homebrew scene forums binary images are supported from ps3 backup managers only on single image binary (i think iso/bin) and works with metadata given by correspondant cue (i think with the same filename of the binary file in the same folder of the cue, and with the correct binary name into the cuefile). Also well known from you all that some games have libcrypt, some have multitrack entries (cd-xa), so there are games that have them both, and i wonder if it's possible to manage them correctly without applying that crappy hashbroker ppf patches for libcrypt.

I downloaded SBITools by Kippykip, a recent tool based on psxtoolz that uses lsd files in combo with cue/bins to generate a redump matching ccd/img/sub/cue:

https://github.com/Kippykip/SBITools/releases

The tool works awesome but again we have img binary and sub subcode files separated. I was wondering if it were possible to combine img and sub into a single file that could be mounted into ps3 games manager. Ps1 binaries have MODE1/2352 format, subcode is 96 format, so i thougth i could mix those two files togheter and obtain a MODE1/2448 binary+subcode that matches correctly the structure of ps1 cds. After some research i found that Alcohol 120% uses mds/mdf in 2448 without cue, so i thought that i could mount ccd/img/sub in alcohol and create a new mds/mdf in 2448 mode with correct settings for PlayStation, so i did it.

Then i checked mdf size and was effectively 2448xNumberOfSectors, so i thought i could copy that file and rename it as .iso, and try to generate a cue with ImgBurn. It detected the 2448 format and created the cuefile (MediEvil (Italy)):

FILE "SCES-01494.iso" BINARY
  TRACK 01 MODE2/2448
    INDEX 01 00:00:00

So to check all the things i used alcohol to see the sectors in search of libcrypt data, but i realised that subcode has a different pattern in iso than in sub (posted sector where first libcrypt sector should be and even is not there (?)):

Alcohol sector viewer
004000000000004000000000000000400000000000000040000000000000404000000000400000000000000000400040000000000000000000000000000040400000004000000000000000000040004040004040000040000040000000000040

ISO
004000000000004000000000000000400000000000000040000000000000404000000000400000000000000000400040000000000000000000000000000040400000004000000000000000000040004040004040000040000040000000000040

SUB
00000000000000000000000041010103080500031005B241000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000

Alcohol sector viewer shows licrypt Q formatted sectors and it matches the 41010103080500031005B241 from the sub, apparently the libcrypt sector seems to be shifted somewhere else than sector 14105. I can't understand how it works, but about a year ago someone on this forum told me that alcohol shifts something, however sector

14105    03:08:05    41 01 01 07 06 05 00 23 08 05 ff b8

is always present but in another location. In fact i found that sector 0 has MSF 00:02:00, so it's shifted of two seconds. why does sbi2sub shifts the msf of 150 frames? in fact the original cd has libcypt shifted in the same way, so i guess sbi2sub is making the right work.

The main thing i can't understand is the way subcode on the disc differs from the content of the sub file. is there a way to scramble img and sub (command line or little executable) into a 2448 iso file as a sequence of 2352,96,2352,96...? Or maybe i have to use alcohol to mount ccd/img/sub and generate a mds/mdf, then delete mds and edit cue to point to iso and force MODE1/2448? I'm getting mad! sad

  • Registered: 2012-09-18
  • Posts: 22

I've found while googling that suchannels are interlaced when burnt on the cds, so the true raw data differs from sub file that is deinterlaced, explaining what i've said in the previous post. Found also a smart tool from themabus that does what I wanted to do, it's called 2352to2448, it scrambles raw 2352 and 96 subs sector to sector (with interleaving or not). So i've been able to create a 2448 image.

I've done many experiments and found that alcohol is able to create such a file starting from a ccd/img/sub set, and results are identical in case of monotrack disc, while in mixed data/audio it eats 150 sectors from the first audio track and generates a mdf file shorter than expected.

Now i have to manage 2448 binary file, cuesheet doesn't seem to support this kind of binary well (at least this says alcohol when burning), even if contains MODE2/2448. What is the correct writing mode that can burn a set of cue/iso?