Extracting simple models

[PaulBuryCZ]

Thank you again shakotay2, you are awesome.

[SchnelleMeyer]

use "File>SaveAs mesh" when you are done, that will save out a working obj file with the settings you have in Hex2obj.
i should note that there is no shortall with Halffloat UV setting, so you have to save out 2 obj files with
both settings and combine the correct UVs (lines that start with "vt") with the correct mesh manually.

Yeah I save the mesh with shortall and UVs with Halffloat and both are good.
How would you combine the correct UVs with the mesh manually as you say? This is my issue that I cant solve.

Thanks.

EDIT: Nevermind my above question - I found a quick and easy way to do this in Blender. Edit no 2: That didnt work after all... Question still stands

[Acewell]

open the obj with correct mesh settings in Notepad++ and press ctrl+h and switch to the "Mark" tab
type "vt " in the "Find what:" field and tick "Bookmark line" and finally click on "Mark All"
now check to make sure only the lines that start with "vt" are bookmarked
now go to "Search > Bookmark > Remove Bookmarked lines"
this file is now ready for the correct UV data to be inserted
just leave it open in Notepad++

open the other obj with with the correct UV settings in Notepad++ and press ctrl+h and switch to the "Mark" tab
type "vt " in the "Find what:" field and tick "Bookmark line" and finally click on "Mark All"
now check to make sure only the lines that start with "vt" are bookmarked
now go to "Search > Bookmark > Remove Unmarked lines"
this should leave only the UV data in the file
ctrl+a to select all and ctrl+c to copy it

now paste it all (ctrl+v) between the vertex and face indices data in the first opened obj file and save it
it should now have all correct mesh and UV data

[TRDaz]

Hi, I've worked with hex2obj before and I can understand the basics but I have come across some files from Mirror's Edge Catalyst where I am unable to find the starts of the faces and the vertices, there seems to be multiple submeshes, I attempted to get it to work but it seems to just output a bunch of messed up triangles. Any chance someone could check it and give me advice? I've seen someone use hex2obj for meshes from this game so I know it is possible, and I believe all the data is in the file I uploaded. After getting the basic idea of one mesh I should be able to do it with other files. Thanks.
https://1drv.ms/u/s!AoEEkLgybKv7hEc9FXciAWt4qR3a

[shakotay2]

where I am unable to find the starts of the faces

WHAT? search for 0000 0100 0200 and you're done.

Why do I have the strange feeling that you're simply too lazy and want to let others do your work?

faith_prolo_upper_body.JPG

So I'll leave the uvs for you.
well, uv pos is 32.

[TRDaz]

Sorry, I wasn't being lazy and making others do the work, generally I scroll through and check for data like that but I think since there was multiple submeshes I got a bit confused, I haven't used this on models with many submeshes before. I just have to find the start of the next submesh now. Thank you anyway.

EDIT: Oh I actually saw the problem I had before, I didn't change it to HF_all and that caused the messed up triangles. Sorry for making you check it.

[SchnelleMeyer]

Thanks for those detailed instructions AceWell. Excactly what I needed

- Also finally had success with transfering UVs from the UVs.obj to the different subs of the main mesh. Problem was I had collapsed the UVs into a single object and that brought the UVs out of alignment with the mesh data.
I had the same problem initally with Notepad++ but it was easy to spot the error with the text and lines.

Last 10 days has been great: Successfully ripping for the first time with Ninja ripper and learning to extract models with H2O.

[Acewell]

I just have to find the start of the next submesh now.

2nd submesh

faith_prologue_upperbody_mesh_chunk.png

if you search for "00 00 01 00 02" you see 4 of these which is a good indication of the number of submeshes.
you will also now know the start address of each submesh face indices and the buffer size from one to the other.
you can input that data in Hex2obj and it will give you the vertex count for that submesh,
you can multiply the vertex count by the stride (40) to get the vertex buffer size and work your way through the vertex block.
but i think there must be a header file some where to go with this chunk file that has all the offsets you need.

[TRDaz]

Thanks for the info! This has helped me a lot.

[cornal]

This .model from konami is a simple model to start to use H2O?

A004_Ojama_Yellow_b.pc.rar

[shakotay2]

This .model from konami is a simple model to start to use H2O?

A004_Ojama_Yellow_b.pc.rar

looks simple to me.
What did you try, where did you get stuck?

[Acewell]

A004_Ojama_Yellow_b.pc.model

A004_Ojama_Yellow_b.pc_model.png

simple yes, but the vertex start address in that sample might throw you off

why do i feel like ive seen this same sample before...deja vu?

[cornal]

A004_Ojama_Yellow_b.pc.model

A004_Ojama_Yellow_b.pc_model.png

simple yes, but the vertex start address in that sample might throw you off

why do i feel like ive seen this same sample before...deja vu?

I found the scrambled alphabet just like in the tutorial, but when i want find the beginning and the end of vertex address, I dont know what to do , maybe i need learn more of Hex signature

[shakotay2]

but when i want find the beginning and the end of vertex address, I dont know what to do , maybe i need learn more of Hex signature

you surely do!

It's a matter of some basic understanding, finding patterns and simple math:

The face indices' (FIs) list is at 0x29160 - 0x2F276, words assumed (it's mostly words, sometimes DWords and, in very rare cases, bytes).
So the size is 0x6116 -> 24854 = 12427*2
Means we have 12427 FIs here.
Enter start and count of FIs in step "go1" and you get the vertex count (3604, lower left list box)

From the picture you can see the vertex stride is 44 so the size of the vertex block is 0x26B70.
(3604 * 44 = 158576 -> 0x26B70)

Ojama_Yellow_pattern.JPG

Subtract this size from the start of FIs' list and you get the vertex start address:
0x29160 - 26B70 = 0x25F0

btw: it's very important to understand that this proceeding and calculations doesn't apply always
because other 3D formats may have additional data between vertex block and face indices block.

simple yes, but the vertex start address in that sample might throw you off

why?

[Acewell]

why?

because of where and how it starts, to the inexperienced this could be confusing.