XeNTaX

I'd like to thank you for giving it another go and from what I can tell its unlocked a number of new models, however a significant number generate a line 48 failed to acquire buffer bytes errors. I'll post a sample but I have the feeling I've pushed my luck with this.

Damn Allods Team.

TheLifeweaver wrote:I'll post a sample but I have the feeling I've pushed my luck with this.

yep, at this point you'll have to edit the script by hand, any change i make to it will break it for the majority of samples.
for your sample change line 39 from this to this because the first pattern does not exist in that file
i could remove that part of the script that tries to draw only the main LOD instead of all,
but you will have to go back to removing overlapping geometry again.

Helloo, anyone still here trying to get the models from this game? here are some examples..


https://www.mediafire.com/file/hutt46jt ... 9.bin/file

https://www.mediafire.com/file/87gkjrdt ... 9.bin/file

Hello again! seems like are some .bin files that had two or more meshes in the same geometry.bin file, that's why it shows error, is any way to separate the meshes and uv maps in these case? because uv + meshes looks so messy. Thanks

https://www.mediafire.com/file/mkp8f943 ... in.7z/file

moonpaladin wrote: ↑Thu Jan 28, 2021 4:43 am Helloo, anyone still here trying to get the models from this game? here are some examples..
https://www.mediafire.com/file/hutt46jt ... 9.bin/file

The Axe's first mesh is simple to get using hex2obj (view link in my sig) while the script seems to look for a lower LOD.
.

I didn't check how to properly divide up into sub meshes:
. (There's also lots of doubles which may need to be handled before.)

It's certainly an odd format. While it's easy to get the data for the vertices, UVs and faces, there doesn't seem to be anything obvious referring to the actual mesh parts. In the case of the MountCar file, there are several tables after the face data, each with 0x44 entries - I can see tables for what looks like bone names, bone name offsets, bone data, possible bone map, and another table mostly consisting of floats.

Is it possible there's some sort of companion file with extra information in?

Maybe you can make some grouping using a byte in the FVFblock:

43e8: 1.404180 1.819670 0.236610 0.346654 0.476176 85 F2 48 FF FF 00 00 00 39 255 255 255
4408: 1.245370 2.008840 0.859255 0.855578 0.430977 53 23 34 FF FF 00 00 00 3 255 255 255 544

50e8: 1.389930 -1.013220 0.811186 0.800173 0.428872 14 3E 6B FF FF 00 00 00 3 255 255 255
5108: 1.587730 -1.342230 0.132218 0.322254 0.475328 68 AF 0C FF FF 00 00 00 0 255 255 255 648

54e8: 1.541570 -1.193150 0.202857 0.346654 0.476176 85 F2 48 FF FF 00 00 00 0 255 255 255
5508: 1.426860 -0.851314 0.435501 0.282457 0.403209 0C 7F B6 FF FF 00 00 00 3 255 255 255 680

9b28: 0.224665 2.833090 1.714700 0.045335 0.982854 80 AC F7 FF FF 00 00 00 3 255 255 255
9b48: 0.146531 0.772728 2.593010 0.309990 0.976342 F0 62 4C FF FF 00 00 00 6 255 255 255 1242

a548: -0.085373 0.458653 2.491380 0.790059 0.232928 80 00 78 FF FF 00 00 00 6 255 255 255
a568: -0.266773 0.961993 3.015780 0.690360 0.330258 7F FF 80 FF FF 00 00 00 30 255 255 255 1323
f8e8: -0.208150 1.265330 2.953150 0.660827 0.211892 5B 80 05 FF FF 00 00 00 30 255 255 255
f908: -0.633336 0.650284 2.545980 0.735946 0.445295 FF 80 7F FF FF 00 00 00 3 255 255 255 1992

Group "30" looks promising (face indices 1323 up to 1991, maybe more):
. Needs further checks.

well, not so bad (but group3 has several parts, biggest here):
. f908: -0.633336 0.650284 2.545980 0.735946 0.445295 FF 80 7F FF FF 00 00 00 3 255 255 255 1992

1c108: -1.215630 0.928266 0.254375 0.047319 0.278180 6A 93 03 FF FF 00 00 00 3 255 255 255
1c128: -1.305700 1.408490 1.460090 0.591820 0.983158 80 7F 00 FF FF 00 00 00 24 255 255 255 3593

(face indices 1992 up to 3592)

Here's the index/group list (so for creating an obj file you'll have to backup this into an array instead of continuous writing of faces in case you want to join "sub groups" otherwise there would be 347 sub meshes):

Left part of pic shows group3, consists of 107 submeshes:
.

wow!Thanks Shakotay. Yes, I assume that they use a list to activate certain geosets of the model and thus form a mount, and then other geosets for another, the problem is to export the mesh with its indices and thus group them and form the exact models of the mounts, although I see that you have been able to achieve it in a certain way, I'll be reviewing it! If you ever get a way to divide them by their indices, that would be great! that game has many good models! I'm gonna try to get more of these kind of models.

https://www.mediafire.com/file/r1h650l1 ... 9.bin/file

Here's the wavefront obj file (no normals):
MountChopperCar.(Geometry)_decomp_0.zip
-----
Here is the H2O file for hex2obj (copy 6 lines into a txt file and name it MountChopperCar.H2O, load decompressed bin, then the H2O file, press 'mesh' button to display; use File\SaveAs mesh to create obj file).

0x75370 101559
Vb1
32 12
0x8 14907
020000
0x0 255

(There will be autocreated submeshes (every 500 faces), though. So better use the obj from the zip.)
Maximum vertex count is 15003, so there's 96 verts left.
.

Thanks alot Shakotay!