IPD : strange format. Don't have a clue on the vertices. When you start reading 378 face indices from 0x 943 as big endian DWORDs (the first two bytes ignored) in APU00FE.IPD there's a max index of 256. But there are too many faces containing only even or only odd indices. So they are very unlikely...

Problem is, I don't know this Mariusz fellow. Where can I find him? Marisuz is szkaradek123, one of my heroes in this forum. Last post to be found here: http://forum.xentax.com/viewtopic.php?f=16&t=7259&p=84226#p84226 (Really don't know why he has 49 thanks only. Should be 4900 or so. :) He...

btw shakotay, is mariusz going to help you do the bones? if you dont mind sharing ur code i can help you get the bones and weights. but i know mariusz is helping you so no biggie if he wants to help you finish it as well. Mariusz send me his SleepingDogs blend file (including python code) when I wa...

Handling triangle strips with version 0.1 With the model c_pnzgren_body1.ps3mesh from this thread http://forum.xentax.com/viewtopic.php?f=16&t=10906 you'll get the head by simply loading the model and the pnzgren.h2o (param) file. You'll have to toggle the 'std' button to 'FFFF' (separator) bevo...

oh, the geometry even before the first terminal is messed up? it's the first "submesh" only. Don't think it's messed up but the normals directions are, imho. Finally I managed to use your code but maybe I did something wrong: http://www7.pic-upload.de/thumb/28.10.13/lm1hhe6ocp3q.jpg edit:...

Thank you very much! :) Just to be sure : I need the boost library to use this? Hopefully it will solve this prob: http://www7.pic-upload.de/thumb/26.10.13/tr3ahqzmjfya.jpg It's the "submesh" before the first FF FF separator ("terminal") where I think the normals marked with a th...

if you are using c++ i can give my strip cut index code. Yep - (it's "normal" C but) that would be very nice. rich though i might have been missing a flag or something but this seems to be my problem, too. i could never find it so i did what only the desperate would do... the vertex norma...

The faces are tri stripped: http://www7.pic-upload.de/thumb/25.10.13/t5zc7667uvy6.jpg http://www7.pic-upload.de/thumb/26.10.13/y6b55pjfjepf.jpg The problem is to replace the FF FF indices. I used the previous index but this does not seem to be correct. edit: hmm, yes, the solution seems to be to sta...

systembest wrote:thx I got mesh data.
But damaged in uv

I know - it's half floats:

phw_00_head_0001.pac.zip

and difficult use calculating hex2obj.

Could you be more specifc, please?
(If you mean the indices count for the head - it's 9510.)

With newest (unreleased) version of hex2obj you'll get this: http://www7.pic-upload.de/thumb/25.10.13/jgaulkbfqtfn.jpg (some errorness faces) edit: ha, stupid me - it's no triangle strips. http://www7.pic-upload.de/thumb/25.10.13/4i5fwe7r8w8q.jpg You can use the old hex2obj version from here: http:/...

Is possible in the future load the entire model and not only submeshes? (for example, the model from dante's inferno) It's planned to introduce a param's file where to store the data of submeshes (vertex start addresses, vertex counts). But this will not release you from searching for the submeshes...

Normals are of subordinate meaning imho because you can autocreate them in nearly every 3D tool - as I wrote. texture coords (uvs) are covered. But your right: the models in this thread http://forum.xentax.com/viewtopic.php?f=16&t=10841&p=89008#p89008 use a vertex block where the uv data is ...

A model from this thread: viewtopic.php?f=10&t=8368&p=89143#p89143



updated exe (big endian untested) in start post.

I made a mesh extractor for simple formats, hex2obj (view link in my sig):

This is a result for one of the models in racengine.ps3:

FAQs Q : how to find the start addresses of vertices and face indices (FIs)? A: as for FIs, search for a "scrambled alphabet" but in general it's like this: I don't know where vertices end and where faces start. The only way is make an assumption, check it. If its wrong, make another one,...