It's ok but i think that the "short all" method must be the one used because you get the normals, how are you converting those 2 bytes in "short all" method? like described in my post before i was thinking that you were using hf kind of conversion but now i see that you were referring to another ty...

I don't see how they could shrink those bytes that was an example only. If I knew what they did I would have told you. :D Humm but how did your program exactly get those normal verts ? It's nothing special, something like xpos = Highbyte*256 + Lowbyte ; (big endian data) To handle signed values I u...

i just need to make sense of it :scaredy: in my experience (in most cases) it takes more time to understand other people's code than to write it from scratch. :D An intermediate method that would make sense to me: set 4 or 5 breakpoints (BP) in the C# project where the vertices, normals, uvs and fa...

then we open an older unity package with UABE that Unity Studio can open and we select a file of the same type (mesh) and click File > Add and select the somename.dat that you export earlier and click File > Save now we open that modified Unity package with Unity Studio and export the compressed me...

export a Raw file from here: Assetbundle-extractor.JPG then try to find point clouds using hex2obj (view link in my sig). Use this H2O file: 0x0 500 Vb1 12 99 0x1384 500 020400 0x0 255 with the noPtC button switched to PtCld . (don't have the time for further investigations. So you're on your own wh...

what does that mean: "does not respond"? Within the last two hours, or what? :D (my philosophy is to completely ignore guys who do not respond :wink: ) Anyways, admitted, I was wrong. Look here for an extractor; I'm just fiddeling around with the raw files it produces: https://7daystodie.com/forums/...

To my knowledge "UnityFS" is a bundle type that noone has extracted so far.

and here's the wavefront obj file of that mirror. Remember to comment out the g gstart_x lines by putting a # in front of them. Then try to put a texture onto wingMirrorL_a_ 3 - good luck. WingMirrrorL_A-Multitest-neu.zip Ah, well, see, scrolling thru this thread I realized that D3FEKT seems to have...

it's all wild guessing atm, the data might be compressed, shrinked, whatever. You could switch to HF_all in hex2obj but after pressing the go2 button you'll see that the result is not senseful: asteroid_HF.JPG v 339 94208 0.067566 or v 5436 0.006973 0.06427 for example are assumed to be wrong vertic...

Thats good but how many verts did you find with your tool? I entered 726 as vertex count. (Normally hex2obj offers the maximum face index (FI) as the vertex count but it can't handle edgecompressed FIs.) to get the floats because in the ones that i did extract i have not found any 47.??? float that...

i'm using a simple java app ("REALLY SIMPLE") to get 2 bytes integer from "index-data-decompressed.block" great ! :) That's what most beginners lack in: to gain basic coding knowledges. So java is a good start! As for your asteroid; I've a deja vue but I can't really recall from my tired mind what ...

guess it's the CAD_ATSV_16.carbin where I have to look at for hierarchy, if any But I'll leave this problem for now (same with the positioning) since uvs is the thing to deal with where I got a little bit stuck atm. wingMirrorL_a is a real little beasty, it has 6 materials and the original uv map of...

this all is a little bit confusing to me (maybe you should give your thoughts more structure? :wink: ) First of all I don't think that it's floats in your asteroid_18800-0x2D60-vertex.block Offset 0 1 2 3 4 5 6 7 8 9 A B C D E F 000000 5D 4C 7D C0 2C 53 30 03 11 FF 74 FF CF 80 7F FF 2ndly when I use...

http://pastebin.com/csQjWwd4 The index list says that there should be 724 verts and the vertex data stuff says that too but when i get this two vertex and faces on only one sample .OBJ i get errors. Need more help with this i think :D f 0/0 1/1 2/2 wavefront obj face indices are 1-based. Add 1 to e...

liuxing wrote:Thanks for your help, this method should look right. But to each file I do not know how to be specific to add the missing part.

I don't see where the problem should be.

Add 4 zero bytes before the "size of first namestring" DWord (and after the last DWord of the index table).