XeNTaX

A BMS script and a Noesis script written for the Xbox 360 version of Marvel Avengers: Battle for Earth.

BMS script for unpacking while Noesis python script for textures handling.

Great thanks to AceWell for his effort on opening textures correctly!

Note that some files are shared by different .xpr archives so if you're asked what action to take when using the same output path overwrite or just skip them.
An Obj Converter can be found here

Bigchillghost wrote:The normal maps always look wrong

could be ATI2 or something

AceWell wrote:could be ATI2 or something

Currently I'm only able to handle some simple DXT1 or DXT5 pixel format images and have little knowledge of other stuffs.
Maybe you want to have a look?
.xpr samples

ok here is a Noesis python script that can open all the textures in your xpr sample files supports dxt1, dxt5, dxt1 packed normal map, and raw
the script mostly consist of converted parts of your bms script and chrrox's xpr Noesis script here
viewtopic.php?f=18&t=8102


a couple of things:
1. if a texture has transparency, the viewport will always display it and a texture may appear invisible
if the alpha covers the whole image, press F11 to toggle viewport transparency.

2. if you export all the textures from the xpr (or any library), the first texture exported is always named the
same of the library that contained the image regardless of the name assigned by the script, i think this is
a long standing quirk of the program and i know no way to correct this either.

AceWell wrote:ok here is a Noesis python script that can open all the textures in your xpr sample files

Um, nice script! I bet I can learn a lot about tex formats from this example then. Thanks!

AceWell wrote:2. if you export all the textures from the xpr (or any library), the first texture exported is always named the
same of the library that contained the image regardless of the name assigned by the script, i think this is
a long standing quirk of the program and i know no way to correct this either.

I've updated the bms script, and simplified your python script to handle directly .bfe files, where I added the 0x40-byte entry to texture data as their headers.
So this problem is solved now.