maybe - but keep in mind that an unknown compression or encryption method might be used.

Only method to reveal this for sure is to debug the game.

(For pc games ollydbg is a good debugger but seems this game was published for consoles only.)

Finding the uv's starting address for char_birds_v2 is a very simple calculation: 0x11403A - 2568x24 (dec.) = 0x104F7A (combined with a courageous step 3 bytes back -> 0x 104F77 :D ) (because there are 3 unknown bytes before 0x11403A) bird_v2_uvs.JPG What I'm a little bit worried about is the holes ...

create a bat or cmd file with this line contained:
call comtype_scan2 D:\test\comtype_scan2.bms D:\test\test.bin D:\test [10000000]

where D:\test is the folder which contains quickbms.exe, comtype_scan.bat, comtype_scan2.bms
and test.bin, the file to be decompressed

Since TextureFinder doesn't reveal anything you could try to use a "comtype_scan" (Quickbms). It will check the inputfile with 500 different decompression algos, iirc. Then you'll need to check the outputted files (<= 500), a tedious task. (But you can ignore all output files with a lower ...

nice script! :) But seems you missed to post lines like these in your last addition: displacement_value *= ORO_scale if (magic_type>=0) then ORO_tfm += transMatrix displacement_value I tried D0.BSO without using ORO_tfm and can confirm your script to work with the pilots. But seems there's still a w...

Rip 1 and Rip 2 is just few meters away from each other though. thany you for the pictures! Sure the rips are showing the same segment of the road? For me it doesn't seem so. Anyway, my thought was to scale and rotate a ripped map. Scaling factor to be obtained from measuring the same object on two...

you' ll need to know the method of encryption for Genki games. (I tried the first 147 algos from quickbms using a comtype_scan at no avail.) For 3DRipperDX, all was ripped fine, but the models are rotated according to camera position, and that made things even harder. Could you show a picture, pleas...

just a quick check: You might try TextureFinder to get bmps from the files.

header pattern?
For example K1.BSO, from address 0x130 you'll find the vertex counts, face indices counts (DWords) with an offset
of 48 bytes to the next submeshes entry.

using hex2obj (view link in my sig):

where would be the fun with hex2obj if it weren't the guessing for the starting addresses?
Same goes for the uvs: btw: your face indices count is too small

I would sincerely appreciate any pointers. :cry: well, 6 weeks of struggeling without result? :cry: I don't have that linkage.bin but from what I know it's about 7 GB of size? Did you ever successfully use offzip on smaller files? If 'no' there are dozens of examples in this forum: http://forum.xen...

yeah, I know.
If I have more time I'll try to finish this but maybe someone else will try since the format is not too hard.
I've the point cloud but the tristripped face indices require a little fiddeling concerning the start of the next submesh.