Extracting simple models

[Shur]

Hi. It does not work v 0.24 for w8.1, requires libgcc_s_dw2-1.dll

[shakotay2]

Hi. It does not work v 0.24 for w8.1, requires libgcc_s_dw2-1.dll

thanks for reporting!

edit: use hex2obj_0.24a from the updated start post
(You'll need hex2obj_0.22.zip, too.)

[Shur]

Thx!! v0.2x: Hex2obj_v0.2x.zip this version works fine

[CZW]

Hi. It does not work v 0.24 for w8.1, requires libgcc_s_dw2-1.dll

thanks for reporting!

edit: use hex2obj_0.24a from the updated start post
(You'll need hex2obj_0.22.zip, too.)

tested and work fine with W7 64 bits

[o0DemonBoy0o]

I'm trying to get models from this file but I can't seem to get the correct UVs.
https://www.dropbox.com/s/43stgga27cal2 ... 0.dat?dl=0
Can you help me out? I'm not sure what I'm doing wrong

Code: Select all

0x44C0E8 1671
Vb1
48 12
0x444EE8 608
120000
0x0 255

[shakotay2]

I have no idea. I've zipped the logged values that should be normals and uvs but seems they're not.
For normals the second column contains values greater 1.0.
And for the last two columns they don't seem to be uvs even when ignoring that most of the floats are negative
the last column doesn't contain values between 0 and 0.5 (or 0 and -0.5).

Maybe scaling the 2nd column's values, logging columns as half floats or shorts, grouping the columns in a different way, whatever, you might get your uvs.

But it's a tedious task and sadly I don't have the time for such.

shak_log section.zip

[lovemarin]

hello! shakotay2 thanks for your tutorial! I have tried to extract models from 300 Heroes, and got dat from .x format files with offzip (according to your posts on thread JUMPX V5.01 (.x 300 Heroes)). The problem is the test.obj always turns out to be messed up, I dont know whether it's packed specially or too complex for hex2obj. Since this game contains so many anime characters, any effective methods of extracting are appreciated. If it's convenient for you to give a demonstration, here is the x file, Nero Claudius or red saber Thank you

[shakotay2]

The problem is the test.obj always turns out to be messed up

it's a matter of the correct params I guess:

002144_neo.JPG

any effective methods of extracting are appreciated.

hex2obj is a helper tool only for a first quick analysis. To be effective you'll need experience or knowledge in maxscript or "C" (for Make_H2O project).

Simple formats like this one (at least for the mesh) follow simple schemes.
You just have to understand the format. There are some examples how to track them in the Make_H2O thread:
viewtopic.php?f=29&t=12756

[lovemarin]

The problem is the test.obj always turns out to be messed up

it's a matter of the correct params I guess:

002144_neo.JPG

any effective methods of extracting are appreciated.

hex2obj is a helper tool only for a first quick analysis. To be effective you'll need experience or knowledge in maxscript or "C" (for Make_H2O project).

Simple formats like this one (at least for the mesh) follow simple schemes.
You just have to understand the format. There are some examples how to track them in the Make_H2O thread:
viewtopic.php?f=29&t=12756

yeah I see, it's a hard work and tedious to extract model and thanks for the tips. by the way, could you please share me the offzip you're using now? All the files extracted with My offzip turn out to be .dat file format

[shakotay2]

offzip.zip

[lovemarin]

offzip.zip

finally!! thank shakotay2, and still got stuck with UV information, let me have a rest first

[shakotay2]

and still got stuck with UV information, let me have a rest first

(uv's preview with v0.24b only):

001244_neo_uvs.JPG

0x19434 = 0x168FC + 2x 12x461 (dec.)
(skipping verts and normals to get uvs starting address)

[luxox18]

thanks for the uv viewer shakotay,very useful for me.

keep your amazing work as always!

[lovemarin]

another question: how to find the start of the face indices list without "00 00 01 00 02", some files are not easy to recognise their start or end address, just like the one below, could you take a look please? Thanks

http://www.uploadmb.com/dw.php?id=1434452744

[shakotay2]

another question: how to find the start of the face indices list without "00 00 01 00 02", some files are not easy to recognise their start or end address, just like the one below, could you take a look please? Thanks

nope, sry, I don't have the time to examine every single sample.
For the FIs' start address just do a trial&error approach.

There are dozens of model samples (of different games) in the forum that I loaded using hex2obj.
Look into the simple ones to gain experience.

(In the end you should learn to understand the models' structure, find the counts which are often preceding the data
blocks and so on.)

it's a hard work and tedious to extract model

it's experience that will ease this, as simple as that