point cloud of uvs looks ok; problem with faces - despite preceding DW counter being skipped
.

0x4150 650
Vb1
16 99
0x248 287
121010
0x2654 24

dunno. All I know is the sample Keezie provided was loadable with Havok tools.

updates for what? Havok?
This thread is dead (since 3.5 years!) for good reasons.

You might inspect the Havok project by PredatorCZ:
viewtopic.php?f=16&t=19691

I also put the viewer in each folder, for checking the model(.dff) and the animation(.anm) Where is that viewer from? Virustotal (Jiangmin scanner) detects Adware Startsurf athg in it. . Pal4_role.exe adware.jpg I know finale00 had already made a script for this old game, but he didn't add bones. Y...

yeah, great! (I was sure the FVFsize to be 136 )

once you have the FVFsize you can get a decent point cloud but I don't have a clue about the face indices (FIs), might be quads?
.

These .csab files cannot be loaded with any version of N3dsCmb viewer. Is there any .zar file with .csab included which CAN be loaded by N3dsCmb viewer with anim played? btw: I got rid of the exception occuring on selecting a .csab anim but the horse still doesn't move. (Playing doesn't seem to be ...

Jamal al dambali wrote: ↑Sat Aug 17, 2019 6:04 pm Hello. Tell me someone how to get uv?

Hello. UVpos is 20, use HF_UV (half floats).

These triangles are basically just 1/2/3, 4/5/6 and so on? In principle: yes. # autocreated linear face indices (wavefront obj notation) f 1/1 2/2 3/3 f 4/4 5/5 6/6 f 7/7 8/8 9/9 ... but this includes texture faces (for "standard" meshes). Since it's a collision mesh here f 1 2 3 f 4 5 6 ...

I have a file that controls the collision on a stage, but the weird thing is that it only contains vertex data. No faces at all. I assume it must be some kind of triangle strips, but I haven't been able to figure it out. I tried auto created "standard" triangles, no strips, and worked: . ...

ALIEN31ITA wrote: ↑Fri Aug 16, 2019 5:32 pmBut wait for get the Dat files i have to use wich kind of script?

ALIEN31ITA wrote: ↑Sun Aug 27, 2017 8:37 pm

(2 years later... )

fmt_pvrtc_pvr.py (Noesis) fails in def parseV3(self):
because self.fmt == 6 is unhandled so far.

but the addresses I've tried lead to really warped faces that only show up on one side of the model. As I wrote: the format is pretty simply - you won't find easier ones: . 00000183-dat.png (For the one "sided thingie": import the obj to blender and invert normals or try out backface cull...

Does this mean the model is corrupted or something? I wouldn't know why. That's what you get using "floats" and an FVFsize of 16 chosen. (View links in my sig to understand what FVFsize means.) Signature "ITOM" at start of the mpb file might be the abbreviation for "MOTION&...