structures:
.

Try out this H2O file:

0x4690 541
Vb1
16 99
0x248 312
121010
0x0 255

assumed to be compressed data - you could try a "brute force" comtype_scan (Aluigi) with 100s of compression types.

see here for example:
viewtopic.php?f=16&t=13222&p=108886&hil ... pe#p108886

This seems a little bit hard for meshes... there should be an easy workaround Guess, there is. Looking at the start of a .dat reveals the block sizes: HeroesAndGenerals-block_sizes.png 0x16: 61f80 size vertex block: 16720 x 24, dec. 0x1A: 149BE size FI block: 42207 x 2, dec. An FVF size of 24 and a...

Read this post: - open .blend file from appended zip in blender 2.49b. - in the leftmost (text/script) window press ctrl-p - import .oct file . ToyStory3-blender 2.49b.png (For Win7 and higher read here: https://forum.xentax.com/viewtopic.php?f=16&t=12153&p=99990&hilit=newgamelib#p99990)

can't serve with .csab - using hex2obj (view link in my sig) for getting a mesh:
.

You need to understand the format of the block at 0x2860, probably uv data: Word log from D:\tmp\00000025.dat ( one line represents 16 bytes in the model file ) Values are shown in decimal notation. startaddress taken from go1 editbox address 0x2860: 0 65292 23986 16896 21490 25600 56345 28160 0 652...

00000025.dat:
0x026E is not start of FIs because 0x0260 is start of vertices.

My guess that FIs might be here <>:
00002860 00 00 FF 0C 5D B2 <42 00> 53 F2 64 00 DC 19 <6E 00>

is wrong, too, because the highest FI (0x25A) can't be found in the block up to 0x4E60.

(I can't access the inked drive in your post as of Thursday.)

sonicdude143 wrote: ↑Fri Aug 02, 2019 4:57 am Does this mean anything?

compressed size of data block -> uncompressed sodb / start of data block I guess

noone will understand what you're talking about if you don't upload the .dat file in question

sonic_hr.mol is compressed - use offzip for example

format looks simple enough - using hex2obj (view links in my sig):
.

The script has been updated since 2014 (surprise ) and I won't dig into this again because the problem with the pmo was solved by patching the old script from 2014.
So guess you're lost for now.

SSJshao wrote: ↑Mon Jul 29, 2019 10:18 pmhex2obj extracts only mesh

correct (and uvs)

SSJshao wrote: ↑Wed Jul 24, 2019 6:13 pmand I can't find an importer of this format for 3ds max.

I can't help with 3ds max - here's what to get with hex2obj (view links in my sig):
.