Extracting simple models

[shakotay2]

Your vertex block 32x16220 exceeds the file size. Lowering the FVFsize to 6 doesn't give a proper mesh, though.

You're better of to search for some point clouds at first.
.

MG_son00_0_gpu.png

[ajspeed11]

Hey can anyone help with this one please.

test.png

I've found Face indices at thiee addresses

3328
3394
3432
36BC

I think these are the addresses of the vertices
3780
37B0
37E0
36BC

I'm not quite sure what i should use for step 2

Any help would be appreciated thanks

https://drive.google.com/open?id=1sHIs9 ... ZkGTSx6QrO

[shakotay2]

I think these are the addresses of the vertices
3780
37B0
37E0
36BC

well, no, that doesn't make any sense to me. Did you have a look at the hex bytes at file offset 0x3780? 192 only and lots of zeroes.
There's many floats in the file (at 0x150C for example) which you don't seem to have recognized.
.

truck_LOD5-tmd.png

(one submesh only)

calculating the face indices count:
0x3370 - 0x3328 = 72 dec. = 36 x 2 -> 36 indices

[ajspeed11]

Hey shakotay2,

I'm still trying to gain an understanding of this process. I thought I'd found vertices at those addresses as there appeared to be a pattern to me at least XD

Thanks for the help. I'll do some more investigating and see if i can progress now you've guided me in the right direction.

Thanks for you help.

[ajspeed11]

Hey shakotay2,

After following your advice I managed to get the meshes from LODs 5,4 and 2

I found that the vertices start after this sequence '010000010000000000000000'
So I did a search for this in the LOD3 and got 695 hits.

I tried the first one which produced this mess, however it does appear to have similar shapes to those from LOD 2.

test.png

I also Looked over LOD 1 and that had 754 hits for the sequence above =/

Any suggestions how I could progress?

Thanks

https://drive.google.com/open?id=1CNzSM ... OrnyYH-zGC

[shakotay2]

Any suggestions how I could progress?

Sooner or later you need to code some lines (see my Make_Obj project)

Or use Noesis (python) for example.
Meanwhile you could search/care for the bigger vertex blocks.

(Offsets between the start offsets of the 010000010000000000000000 sequences vary, it's 0x80, 0x100 or greater.)
.

rtruck_LOD1-ea3.png

.
For LOD3 you could search for 000001000200, there's 45 findings. Get end address and max FI count of each FI block, then try to find suiting vertex blocks. (FI = face indices)

[ajspeed11]

Hey shakotay2,

Thanks for the fast response and for taken the time to look over these files again.

[Hell Inspector]

I was just wondering if you could possibly help me with this model from Wallace & Gromit TCOW cause I don't know anything about hex stuff and maybe you can like teach me certain stuff about it. https://cdn.discordapp.com/attachments/ ... tor_US.ovl

[shakotay2]

I was just wondering if you could possibly help me with this model from Wallace & Gromit TCOW cause I don't know anything about hex stuff and maybe you can like teach me certain stuff about it.

well, I'm very busy with my own projects. I wrote a tutorial (click tut button in hex2obj) to help people help themselves.
(From a quick glance I'd say this .ovl contains animation data.)

[Mediocrity]

What about this. It's a cylinder with 24 vertices from toy story 2 labeled "paint" but I don't see the face indices.

paintfile2.png

I tried the fake option but no success unfortunately. By the way, where could I find more info on indices. I can get point clouds mostly on some games but I don't know how to identify indices that don't look at least similar to the ones used in the tutorial.
Here is the file if you need it: https://drive.google.com/open?id=14YsgD ... xaSOdJszqA
Thanks.

[shakotay2]

I don't even see a cylinder.
What I get is this:

Code: Select all

# 0x4a: verts= 24
v 127.000000 -66.000000 119.996094 
v 66.195313 49.230469 67.585938 
v 127.000000 66.996094 119.996094 
v 67.195313 49.230469 67.585938 
v 127.000000 66.996094 119.996094 
v 67.195313 49.226563 66.585938 
v 127.000000 -66.000000 119.996094 
v 67.195313 49.230469 67.585938 
v -20.007813 56.996094 97.617188 
v -100.000000 -126.441406 -99.804688 
v 64.503906 -56.000000 0.000000 
v 94.503906 43.000000 0.000000 
v 0.000000 -56.000000 127.246094 
v 119.996094 -66.000000 127.125000 
v 0.000000 -56.000000 127.246094 
v 119.996094 -66.000000 127.125000 
v 0.000000 56.996094 127.246094 
v 119.996094 66.996094 127.125000 
v 0.000000 56.996094 127.246094 
v 119.996094 66.996094 127.125000 
v 52.070313 4.000000 -56.000000 
v 119.996094 -66.000000 0.000000 
v 52.070313 4.000000 -110.996094 
v 0.000000 -56.000000 0.000000 

[Mediocrity]

Well...I'm stumped. I can't seem to get the values you posted. This is what I'm looking at.

Screenshot (315).png

I can get point clouds of the other models using similar settings but that's as far as I get. Thanks for looking at it. I'll keep trying.

[shakotay2]

Well...I'm stumped. I can't seem to get the values you posted.

sorry, I used the box.all file with big endian.

This is what I'm looking at.

I see. Could be a matter of using another tri strips algorithm for automatic creation of face indices.

Current solution is this (f a b c, f b c+1 c):
# autocreated tristripped face indices
f 1 2 3
f 2 4 3
f 3 4 5
f 4 6 5
f 5 6 7
f 6 8 7
f 7 8 9
f 8 10 9
f 9 10 11
f 10 12 11
f 11 12 13
f 12 14 13
f 13 14 15
f 14 16 15
f 15 16 17
f 16 18 17
f 17 18 19
f 18 20 19
f 19 20 21
f 20 22 21
f 21 22 23
f 22 24 23

[Mediocrity]

I was wrong. There are 48 vertices. There's a pattern of 4 vertices followed by 16 bytes of other data and the pattern repeats. I guess the vertices are grouped per face. Maybe that's why there's no face index. Sounds like it can be done by someone with the knowledge (me?) if the pattern is consistent in the other files.

[shakotay2]

well, since uploadmb is down I'll create a new (temporary) hosting here until I've found another reliable hoster.
Just unpack the three zips (part 2 and three see following posts) into one directory of your choice.
-------------------------------------------------------------------------------------------------------------
btw: SecureAge APEX says the zipped exe here has a virus - I have no idea, why.
66 other virus scanners say: no virus found.
So if you don't trust it don't use it - as simple as this.
-------------------------------------------------------------------------------------------------------------
.

Hex2obj_0.24e3-rel,1.zip