This topic was archived here:

https://web.archive.org/web/20230000000000fw_/https://www.zenhax.com/viewtopic.php?t=545

AlphaTwentyThree, posted Mon Jan 19, 2015 5:20 pm (2752)

---

Hi there!
I wanted to look at the music from the above game and stumbled across an audio codec that I've never seen before. The byte distribution reminds of uncompressed PCM but the structure is completely different. There's one big file without TOC but I'd like to play around with a header if available.
Here are the first 16mb of the archive: http://*USE_ANOTHER_FILEHOSTING*/0710fe8 ... m.dat_16mb
Has anybody seen this codec before? Is it maybe some kind of obfuscated PCM?
Any help is appreciated!

aluigi, posted Mon Jan 19, 2015 11:51 pm (2767)

---

Code:

filexor "\x8e\x07\xfa\xe9\x8c\xdf\x8e\xf1\x83\x1e\xe1\xbf"
math OFFSET = 0x1c
get SIZE asize
math SIZE -= OFFSET
get NAME basename
string NAME  = ".wav"
log NAME OFFSET SIZE

AlphaTwentyThree, posted Mon Jan 19, 2015 11:56 pm (2768)

---

damn, such an easy solution... THANKS LUIGI!

AlphaTwentyThree, posted Tue Jan 20, 2015 12:20 am (2773)

---

Well uhm... I'm having problems with writing a correct script for extraction of the archive... Could you eventually take a quick look for a short script? That would be great! Here's the complete archive: http://*USE_ANOTHER_FILEHOSTING*/5c714ff ... 43/bgm.dat

aluigi, posted Tue Jan 20, 2015 12:47 am (2778)

---

http://aluigi.org/papers/bms/others/qp_ ... gerous.bms

AlphaTwentyThree, posted Tue Jan 20, 2015 1:10 am (2779)

---

You're my hero, Luigi!

aluigi, posted Wed Feb 03, 2016 4:11 pm (11191)

---

I have updated the script to version 0.2 but there are no good news.
Those files that look "strange" are compressed but I don't know what algorithm or customization it is.
So now the files have a correct filename (they were wrong before) and are saved "as-is" if they are compressed.

My guess is that they use a modified lzss, maybe just some parameters.

AlphaTwentyThree, posted Wed Feb 03, 2016 7:53 pm (11196)

---

Interesting news indeed. Maybe something that could be implemented when somebody has the nerve to fiddle around with the parameters.
Thanks for the update!

id-daemon, posted Sun Feb 07, 2016 1:29 pm (11286)

---

I think it would be easier to find the algorithm by debugging.

p.s. just started debugging and immediately found what that was.

the program to unpack here: https://github.com/henteko/2012TeResAI/ ... /DXArchive

key = qppassqppass

aluigi, posted Sun Feb 07, 2016 7:20 pm (11294)

---

Well done.

Just a note, I guess that the official homepage with the updated tool is http://homepage2.nifty.com/natupaji/DxLib/dxdload.html