I have almost uncompressed this, but I haven't been able to figure out this completely. (Anything is little endian). Header 4 bytes: "Wb16" ASCII string 4 bytes: Uncompresed file size Custom-LZSS Compressed blocks (Come immediately after the header). 4 bytes: Compression flags. 32 * 2 bytes: Compres...

The format looks like a expanded version of Nintendo DS WAV (SWAV). http://kiwi.ds.googlepages.com/sdat.html#swav Big endian byte order. RWAV (first) block: 4 bytes - "RWAV" (Revolution WAV?) 4 bytes - Magic (0xFEFF0102). 4 bytes - Filesize 4 bytes - Size of this structure 4 bytes - Number of blocks...

Yes, it seems to be VOX ADPCM, and the header is 512 bytes (pretty simple, I've documented it in ScoreHero).

With sox, after removing the first 512 bytes, and decoding from VOX ADPCM, it sounds pretty good, there are some imperfections, but I guess that those are because of the low audio quality.

I think that there is no safe way except formats... every combination of audio values is valid, so it's impossible to do it with a 100% precision.

Also, a lot of games don't use PCM... in X360, there's XMA, and in PS2, there's XA ADPCM/VAG.

You can also try with some format detector tool...

EDIT: I didn't see the sample. Looking at it now EDIT 2: Nothing. It wasn't it. I converted the sample manually right Probably, you've started to get the palette from a wrong place. It starts at offset 272, it's 1024 bytes long. The image seems to have blue and red colors swapped according to the "o...

It doesn't sound like a known format, probably it's a format specific from a game/company

In that case, something must be designed especially for it.

I have no idea about how to convert the BND entries to TGA (they're images, right? I will seem completely retarded if they aren't :roll: ). Here's what I have now :oops: struct bnd_image_header { int unknown1; int unknown2; int unknown3; int size_image_data; int unknown4; int entries_palette; } __at...

Well, finally solved it (thanks to Wikipedia: http://en.wikipedia.org/wiki/Highcolor ) It's just 5 bits for red, 6 for green and 5 for blue. Here's my swapping formula. Notice that palette[] is an array of integers, so I'm swapping 2 colors at the same time. palette[i] = (palette[i] & 0x07E007E0) | ...

No. It's a 16-bit (color depth) palette (this means, 2 bytes per entry/color), with 256 entries (with a total of 2*256=512 bytes). I just want to know, of each entry, what bits are green, what are red, what blue, and what alpha (not sure if there's alpha). For example, in a 32-bit palette the colors...

I'm trying to convert the GH3 image files but I have a little problem with some files with a 16-bit palette. Apparently they have the palette messed up, in 32-bit, the order is "BGRA" instead of the normal "RGBA". I can view the 16-bit images now, but I can't figure out how to change it. I've tried ...

I'm not an expert in graphics, but the "bkn" header seems to be: (char[4]) "BKND" (magic) (int) version (int) number of entries (40 bytes * number of entries) entries Each entry (int) filesize (int) offset (32 bytes) identifier? No idea about the files inside it. I think that there's a palette :(

Here's an small unpacker (in C). The file names are just obtained from the 4 first characters of every file. I think that the section after the file names is encrypted, too. If you go to the the part after it, there's something like: (8 byte) ??? (int) Size of something (probably the filename) (File...

Here there are the basic file details about PKD files. They're used to store files (probably, it means PacKeD)... File header: (all field are little endian) (char[4]) magic "PKDF" (int) reserved? (int) version? (int) number of entries (int) size of the file table - 4 (not sure what it's used for) (i...

Are you sure that they're exactly the same file?

I haven't looked at it too much, but both files seem to have a unencrypted header, at least. The data doesn't look encrypted anyway. It seems an image.

john_doe wrote:Yes, that would be good.
I checked, and the bytes used to XOR against are different for each file. I guess it's calculated via one or more values from the file header.

EDIT: Solved