dunno, looks like a collection of heads and hands, there's 160 assumed submeshes, many of them with less than 20 indices.

hand?

0x85149 605
Vb1
43 28
0x827F5 246
020000
0x0 255

But how should I exactly identify the correct start address of the face indices ? This is the code for *. ps3 files (pFBuf: pointer to file buffer): pFBuf= pTmp ; pFBuf += addr[cnt+1]-24 ; j = addr[cnt+1]-24 ; // use next address [cnt+1] while (*pFBuf==0){ pFBuf--; j-- ; } addrFI = j - 2*dwFaceIndC...

Roman wrote:And i get all objects in same places like i see in 3d level unreal 3d editor.

Need structure Unreal Engine 4 data level.

can't tell you about UE4 but do a forum search for 'parent'.

In the finds (297 so far) you might search for matrix, matrices or multiply.

HTH

Is it possible to write a tool with hex2obj built in to extract them automatically? If only I know what to do about it. Such tool can be created using this project: http://forum.xentax.com/viewtopic.php?f=29&t=12756 It requires that you do a format analysis and have some C coding skills. There's so...

Looks like double faces (maybe foreside and backside?). Swapping means the winding is changed. You can check such problems in a very simple way: create a copy of the whole wavefront obj file. In the original one delete all odd faces (line 1, 3, and so on in the faces' list). In the copy delete all e...

face indices start address is correct; but it's 5 submeshes, search for 000001000200 to find the different startaddresses next find: 0xFF04 (delta 0x2FF4 related to 0xCF10-> delta/2= 6138 face indices; vertex count= 1180) Why set start of vertices to 0x1387? That's not senseful. calculation CF10 + 3...

uvs not found; maybe uvpos 24 and 32? (hex2obj uses uvpos, uvpos+4 for floats).

My most preferred one is WinHex. But any other should do which has a decent search feature.

For me a byte grouping of 8 plus 8 is totally important (which HxD v 1.7.70 for example lacks of).

But when messing with other chunk files I was trying to figure our how you got: "5CCE0: start of face indices (FI) list" Assumed first face to consist of 0, 1, 2 (-> 1, 2, 3 in one based wavefront format) so search for 000001000200 (little endian of 0000 0001 0002). Doesn't work always because the ...

@eatrawmeat391: dunno whether you got the uv data so far; seems he's reading it kinda this from 02.yobj (2 floats per address, in address steps of 0x20): 0, 0 0x2868L, 5312DC3D, 00482B3F 0, 1 0x2888L 0, 2 0x28A8L ... 0, 15 0x2a48L 1, 0 0x2a88L ... 1, 15 0x2c68L 2, 0 0x2ca8L 2, 15 0x2e88L 3, 0 0x2ec8...

@eatrawmeat391: nice work! :) But it's very unlikely that someone can test it if you don't explain, how. Which platform? setup.py crashes blender 2.49 (python 2.6.2) on windows. When starting yobj.py (ctrl.p) it doesn't find your modules (control.py, data_op.py, same directory) copied all your modul...

Hi everybody, there's a way to open the .sb3d file models from this game :P so, what is the way? From a quick glance at cop_ford_explorer.3b3d it looks like there's face indices, a hierarchy part with submesh names, and an assumed matrices part. For the mesh part, well, maybe compressed. There's no...

May I ask why you are opening another thread on IOS .sb3d, NFS no limits?

viewtopic.php?f=16&t=15433&hilit=sb3d

yep. @eatrawmeat391: "For every vertex (triangle point) in the model, we have to figure out which bones it should be attached to, and how strong each attachment should be." read here for example: http://blog.wolfire.com/2009/11/volumetric-heat-diffusion-skinning There may be up to 4 weights per vert...

chunk files; continuing from this thread: http://forum.xentax.com/viewtopic.php?f=16&t=13675&start=45 @ekmek: handling the chunk file I show the steps which should be plain to see for any noob; if you don't, well, then I fear I can't help. address 0x5CCE0: start of face indices (FI) list search for ...