well. for now here are all the wheel textures i could harvest. includes what seems to be lod textures too. automation is impossible. the palettes can not be identified and relation not be done. and the palettes use those triple surfaces which i had to manually split to not complicate things. https:/...

i might be a lil rusty, but we gettin' somewhere with the code. i need a break now tho. brain drained over nite. and i got a stream to watch.

well looks like i found some other form of palette in the wheels file right now. will continue and start coding then. this gonna be a mash of bms and noesis. i dunno how to export multi image files from a noesis script. things to learn. i need the ps2 swizzle from noesis tho. palette for the rim, th...

are you serious? the palettes and texture have no relations. how do you think this gonna work? i sure can read the hex. this is a hard egg i think. ps2 data is kinda always a mess tho.

Acewell wrote: ↑

Mon May 20, 2019 12:17 am

episoder wrote: ↑

Wed May 15, 2019 11:36 pm

maybe somebody wants to recode this lil script to py syntax.

what is the game name and platform from which the samples came?

actually unknown. hence why i msc-ed it. dude removed it anyway. i think theres no need to convert anymore. you wanted to do it? yo.

yep. and it was not about it being easier, but the batch process. qbms doesn't export png either way. just figured noesis has a batch function too. i never tried it tho. not gonna do now either. maybe somebody wants to recode this lil script to py syntax. i dunno. and dds and the old dxt formats are...

oof. looks a bit messy. seems hard to get the structure data pointers. 2 surface descriptions. those are the 50,51,52,53 bits. it's actually only 1 tho. 5 kinda 'fragment' descriptions. but... nasty to figure out the positions and sizes. the leftmost textures are 2 fragments to puzzle together. tric...

you know quickbms? should be able to batch it, safely. formats afaifigureditout. :D ask if you need more. gotta up the 'broken' samples ofc. i'm somewhat certain it could be done in noesis too, but it's a execute "import and export on click file in folder thing" unpredictable reads and writes withou...

you basicly reassemble the index list to index the per triangle vertices. this is clockwise order.

try https://github.com/Perfare/AssetStudio and open the whole folder. AS usually combines those split files into full files and is able to open the whole thing. depends on the file's hierarchy and your choice if you wanna extract from the scene or the asset list.

you know it's just the first mesh in this file tho, do you?!? anyway. glad you figured it out.

at 0x430 you have the offset and count for the index buffer. this probalby belongs into this $RSI structure. guessing the "I" is something index releated.

interesting sample. you interpretation is not entirely correct or imcomplete. anyway... if you know the file that contains the buffer lenghts or vertex or index counts or whatever describes this data container you should load that first and then the data file. how you do it depends if the descriptio...

i throw it. :| been working on this all day. blender doesn't give out correct raw triangles. neither correct colors. the indices are all over the place. i got correct colors when splitting (really splitting) the polys/triangles on a test quad. pretty blah. it may work when you get to seperate all po...

i'm messing with blender py rn. maybe i'll have an exporter done later. or maybe in a couple hours (tomorrow).

(i need to learn it anyway. gonna need an exporter for my own engine model format at a later point)