We should refrain from hijacking wobble's thread; see my answer here:
viewtopic.php?f=29&t=14695&p=124145#p124145

btw: your uploaded pics show that you don't seem to have read the tutorial (tut button)

(in case you don't need weights and bones:)
once you've got it as obj there should be little problems to convert it to fbx

the position is handled in v section. I'd suggest to refrain from wild guessing and use a textured square in blender (not a cube) to understand how things work. (press 'n' in blender's 3D view, in properties panel scroll down to 'normals' and press button "display face normals as lines") Export he s...

looks like y and z are swapped

from "r_eye" and "l_eye" it looks to me as if it were a skeleton file with matrices.
(too busy to care for now)

It's not the only way, for sure, but it's the simplest way I could think of
in case you can't script or code and don't have the luck to be a UU3D user.

you could try it for yourself: (There's also an FVFsize of 40 to be found in the bin file.)

As you can see from the "resulting listener output" it worked.

You've to change the path (K:\\...) so that it points to your model files' folder, of course and set the model_type.

Thank you very much!
Great to see your progress; seems you're walking a steep learning curve.

I remember the times when this used to be a "research forum".
What were your investigations so far?

With the hint from finale00 it's plain to see that there's a png from 0xC0B0 to 0x19406 in em016.pac: btw: aluigi wrote on Zenhax that files containing 'ecd' are probably encrypted.

yeah, scripts! As you may know a full format analysis is required before you can create a script. We've some capable "scripting heroes" in this forums but they can't/they won't care for any format as you may have experienced. A last time (hopefully) I'll show how easy it is to obtain the model by yo...

the patch is simple, replace elif parVer == 259:
by

that's great!

Could you please give a brief survey about the main points to get your solution?
Or your final code.

Because there was transposing, multiplying and inversing matrices.

Transposing was only required to adapt the opengl matrices for DirectX, right?

there's no simple adaption for the script, afaics
Meanwhile you could fiddle around with hex2obj (pl00_head.msb):

Everything you need is right there in front of you. yeah, I see. :) But I'm too busy with my own projects than to waste lifetime with annoying formats such as "nif". And as you wrote, that format is "solved", more or less. I found the NifConvert project here on my old pc (guess from skyfox69 , http...