Still toying with the maxscript and had one file with an issue so far that I sent you a link for tha didn't work well with the script but not sure why. I've PMed you a quick hack for the HelmetOfTomorrow . (btw: forget to mention this unk var to be used now: delta = readlong stream -- 03 000000 ? f...

So I'm guessing this maxscript will allow a person to import the .dat into 3ds max and export as a obj from that? check your PM Guessing this is what you got as well when running the numbers too for the Vex Mythoclast weapon? Not sure why it looks like a shield but that's better than what I had whe...

but the one you posted looks strange, was that with the maxscript? yep. In blender it looked good to me - seems that damned blender obj importer repairs everything magically. :D AgemaType3_b.jpg (ok, smoothing doesn't work.) But there's no superfluous faces anymore - is it? It's like every other fa...

Maybe with the Max script it will make doing these models much much easier. With hex2obj my fault was putting the handling of normal tristrips and tristrips with FFFF end marker into one function. Together with the endian handling and some other checkings (don't know anymore why I introduced them) ...

0x1A105 is simply the wrong start address for the face indices - using 0x1A103 improves things

[...]but I will study this tool because I think each mesh uses diferent values, I don't know how to set them. Seem so hard! You'll need some basic understanding, that's what the tutorial (button tut) should be good for. Once you've done the example in the tutorial things should become more clear. T...

well, with blender the helmet is displayed properly. If you meet an outside-in behaviour of the mesh it normally can be fixed by inverting the normals (edit mode, mesh tools, flip normals in blender 2.49b). Another caveat might be the tristripss algo used in hex2obj which is creating too many faces ...

using hex2obj (view link in my sig): (with UVpos = 8 the UVs look a little bit weird, dunno why)

Shakotay I think I love you. Love you, too, now for not giving up. :) btw: offzip is a genius program from genius aluigi, so don't damn it. :D . The directory thingie is a little bit confusing because the usage hint says: "<output/dir>" whilst the console log says: "in directory"...

copy offzip.exe and your geom into a folder, say I:\xxx, then I:\xxx> offzip -a 2994845057.geom I:\xxx 0 you should get 00000004.dat in I:\xxx console output: Offset file unzipper 0.3.3 by Luigi Auriemma e-mail: aluigi@autistici.org web: aluigi.org - open input file: 2994845057.geom - enter in direc...

did some translations: puzzleMe.jpg (Looks like an improvement but sadly is far from being correct.) This is the skeletal hierarchy: # 0 sanjuro # NumChildren: 1 1 pelvis # NumChildren: 4 2 right_legu_1 # NumChildren: 1 3 right_legu_2 # NumChildren: 1 4 right_legl_1 # NumChildren: 1 5 right_legl_foo...

I looked at some noone lives forever ltas which seems to be the old format that won't help with ABC format, I guess.

thx. They possibly could if :) - you had an lta file that was derived from an original ABC v6 file. The problem is how to calculate the positions. The ABC v6 format specs says: the vertex positions are "relative to object's origin". I was thinking to add the parents pos as an offset to the...

here a test version (v0.0) todo:

• calculate absolute positions
• uvs

Note that sanjuro has 694 vertices but maximum face index is 390. Dunno why.

first step done - I succeeded in separating the submeshes: separateSMs.jpg Just tested it with sanjuro giving decent submeshes, too, by the sorting of the faces, very strange, since it's the same vertices. Also half of the vertices are unused so the char is uncomplete. I'll upload the exporter (mesh...