bin2seq usage by loxrox at 9:02 AM EST on February 23, 2023
Hi I was trying to extract the music for the ps1 game called Goujin Senki. The VB and VH were easily extractable, however seq was stored as SEQDATA.KSP. I tried with vgmtrans and vgmtoolbox to extract the seq, but with no real success. I feel like I might have some luck with bin2seq, however it feels like I am doing something wrong with the command syntax. I used: bin2seq -a (path to file) 0 The issue is, that it doesnt give you any sort of feedback, if something is incorrect. I tried to use other game files mentioned in the forums, that worked with the program, but I could never make it work.
Programs do not give "expected" feedback because... How do you TELL that program what "expected" format of INPUT file it expects? File extension is mostly MEANINGLESS when extracting packed data; one should examine contents in order to determine if input file can/could be unpacked/extracted. Program cannot "know" beforehand, it's mostly a manual process of examination.
AT least, please upload this "SEQDATA.KSP". Data chunks inside it MUST conform to "SEQ" format before being usable. Manual splitting ought to be expected.
How did you extract "SEQDATA.KSP"? It would take me time to convert 0x930 (MODE2, "RAW") sectors to 0x800 (MODE1 / ISO) ones. If possible, please re-extract file by A) copying file from image/CD drive to PC, or B) extracting ISO contents in a mode different than RAW. If calculation is OK, this same file in ISO mode should be 754 kB in size (instead of 865 kB).
Comprissed of "chunks" each one starting with more-or-less 0x100 bytes of almost [0x0], part of a TOC of 0x400 bytes size (the "header" of the chunk), with relative offsets to very many chunks that start with "KSQ." [4B 53 51 1A]. Theose headers have [00 04 00 00 00 00 00 00] (including header size) qwords, so using this as search parameter I found 38 of these chunks having on average about 64 "KSQ."s (subchunks). 2454 "KSQ."s found, some with data (large, significant string of bytes), some others without.
Apparently there is much "parity" (symmetries?) on the bytes/dwords, many [FE] bytes on average; data structure reminds of SEP (container of SEQs), data seems obfuscated/cyphered rather than compressed. Might be a custom SEQ container.
Oh well I suspected as much. I probably have to record everything ingame then. I already checked for some sort of soundtest to no avail. There are at least some cheats so I can get through the game quicker.