what about the files which start with the sequence of bytes "0f f6 12" ?
for example that music_church_combat1_lp.spk file which was linked some posts before

job done:
*edit* watch the link in the next posts

in case someone needs it, this is the quickbms script:

this is the same format used also in other games like Wolfenstein and the game posted here viewtopic.php?f=21&t=3891

don't use rapidshare, it never works:
viewtopic.php?f=17&t=3792

the extraction of the files is trivial, the problem is their decompression because I don't know what algorithm is used. I only know that it's the same/similar header used in Wolfenstein and other games. in the meantime this is the quickbms script for the extraction of the files "as is" (so...

scanned with quickbms from offset 0x15 and found nothing (maybe the offset I used was wrong)

http://aluigi.org/papers/bms/inquisitor.bms
remember to use quickbms 0.3.9 because I implemented a easier way to handle the chunks of compressed data compressed with slf_block

I have written a bms script for the dgp files:
http://aluigi.org/papers/bms/defense_grid.bms

you must select only the file contains the number 0000 like data0000.dgp or enus0000.dgp (so the first dgp of the collection) because that one is the index file

I guess this topic should be a post to the thread viewtopic.php?f=10&t=3692

1) an output file of 0 bytes means that the algorithm is not the good one 2) quickbms doesn't create the output files for the algorithms that have failed (they return an error if the compressed data is invalid), so they are not created just because doesn't exist a decompressed data obtained from the...

for example in the last file you uploaded we have something like the compressed size (32bit) at offset 0x82, the uncompressed size at 0x86 and the raw data at 0x8a. I have already scanned that raw data with quickbms without luck, and I have also cut the first 4 bytes of the raw data as additional ch...

the only recognizable thing I see in the code you posted is a part of an encryption algorithm commonly called MT_decrypt or MT. while in the testfile is not clear where starts exactly the raw compressed data. in any case I guess you will find the following post enough interesting: http://forum.xenta...

I named lzo1x because the nrv algorithm (also known as ucl) is created by the same author and while I have never found a game/software implementing nrv there are tons of others that instead implement lzo1x.
so only a calculation of probabilities :)

it's simple:
- copy that code in a text file, for example adb.bms
- download quickbms: http://aluigi.org/papers.htm#quickbms
- double click on quickbms.exe
- select adb.bms
- select your adb file
- select the folder where you want to extract the files
- ...extraction in progress...
simple :)