XeNTaX

Hello everyone,

I have been trying for the past 3 days to extract those files, without success regarding models and levels.

I've downloaded around 12 versions of MGS, Gamecube, PC, Playstation...

and no matter what .DAT file I open with Konami DAT utility, it fails,statting that "This is not a compatible dat file"


Then I've tried to use GzsTool to open thoses .DAT file, but here again it fails, Command Prompt closes as soon as it opens and no files are produce during this milisecond.

So I cloned the Git repository, built my own .exe tool and same thing goes.

Even tried with every release of the repository that existed but same result.

In despair I started to recreate from scratch that level where Meryl gets shot using Maya, but texturing is a mess it would be so much easier otherwise.



Can anyone help ?

Thank you

one or two .dat samples?

(In case it's a matter of a changed "file encyption key" you're lost, I fear.)

Well apparently many people seems to have succeeded, I hope it is still possible

FACE.DAT
BRF.DAT
RADIO.DAT
https://gofile.io/?c=YQ78iQ

Thanks! Well, seems it reads a filecount at offset 0x00000008, xored with xorMask2 = 0x11C22050.
Gives a funny high number!

First 8 (or 16) bytes of your .dat files don't seem to have the same structure - so not surprising that the tool can't handle them.

Seems you need any .dat file that can be handled by the GZStool.
Only then there's a chance to compare to your uploaded .dat files.

Sseast wrote: ↑

Sat Apr 25, 2020 6:22 pm

Well apparently many people seems to have succeeded, I hope it is still possible

Any chance to get a .dat file they used?

Then I'll try to put my hands on other .DAT files.

shakotay2 wrote: ↑

Sat Apr 25, 2020 7:01 pm

Any chance to get a .dat file they used?

Unfortunately many forums links are cleaned or expired by now, I'll have to keep digging

Thanks for your help.

If anyone has heard about another .dat file opener or method, don't hesitate to reach me, I'd love to make a diorama out of mgs, I did it for halflife, such a pleasure

I finally suceeded , I stopped trying with the .dat file, instead I tried with the .mgz file created after MGS setup.



Inside, there were .kmd files that I then converted to .obj using http://secaproject.com/kmd2obj_2_51

I also found the textures files contained inside which were .pcx that I batch converted to .png using IrfanView (.pcx files where contained inside a .dar file, opened using unmassw_092.exe).

But the textures name don't seem to match and I end up with untextured model

On one hand, .pcx textures files name are very explicit and seem normal :
- m03_wc_flo1 (m probably stands for material, 03 for level 3 then wc floor.
- m3r_path_door1
- m3r_tu_wall2
...

But the .obj textures filenames looks jamed, I am not sure if it is caused by the converter but there is definitly something off : usemtl Material_00a_2018 doesn't ressemble to the naming system by .pcx texture files

So before I try to mannually repath over 300 textures, did anyone ever encounter this kind of issue ?

Here's an upload if someone is up to investigate
http://www.mediafire.com/folder/mbmpa8l ... 6r9/shared

Edit :

Even the .mtl file refers to an unkwnown naming system/file :

Hello Sseast,
If you wait a bit before he release his tool, my friend Secaproject works on MGS in Blender:
https://www.youtube.com/watch?v=zZ5XsHtazGY
The textures "names" are hashs, the real filenames are only on the PC version (MGSI).
Example: "m03_wc_flo1" = "8206"

Your friend is a bless, I love the tools he developped

Well I think I'll be finished any time soon, but I'd definitely take a look for other levels I'd like to work onto.

I am actually working through a mgsi folder, I've exctracted both textures/models using stage.mgz as I have no success with .dat

How did you know that 8206 = to m03_wc_flo1 ?

If you have a hash table to share, I'd be interested to have it too as I am repathing a little blindly

Also, I noticed that there were polygons missing sometimes, not sure what is causing it but it requires doing some poly reconstruction and UV mapping, not so fantastic