PAA File Format

PAA texture file structure

Introduction

Of the many image file formats out there, such as jpeg, or gif, Bohemia Interactive choose to use a specially developed file format (paa) as the base texture file for all engine types.

The reason for this is the raw data within the file contain mipmaps which be passed directly to Microsoft's Direct X as a DXT1 picture (eg) without further massaging.

All engines except Elite also support JPG files (but they have no transparency). Most of encoded mipmap formats in pax files do.

PAC files

.pac files are (almost) synonomous with .paa. In the origins of OFP cwc. paa were meant to be two tone colors (black & white or grayscale) and paC = colour.

The distinction does not exist. All engines treat pac or paa equally. They are referred to as paX files in the rest of this document.

Note however that Arma2's tgaviewer *cannot* read a palete index .paa extension. only .pac extensions. (the contents can be identical)

OFP DEMO files

The initial 'proof of concept' for ofp released in 1997, was a demonstration of an island with models and textures.

These pax files can still be read by tgaviewer! (and pal2pace)

The *only* difference between these, and subsequent pax files is that the index pallet appears at beginning of file. In other words

• There is NO header type, Pallet index is assumed.
• There are NO taggs.
• Other than this, demo files are 100% identical to ofp's Pallet index including the use of run length compression.

Main Format

Overall structure of a pax file is

struct overall
{
 ushort   TypeOfPaX;        //OPTIONAL
 Tagg     Taggs[...];       //OPTIONAL
 Palette  Palette[...];
 MipMap   MipMaps[...];
 ushort   Always0;
};

TypeOfPaX (optional)

With the sole exception of OFP index palettes and demo, all paX files begin with a 2 byte 'type' signature. OFP index palettes have no type, they begin with tagg structures immediately. demo files have neither, and begin with the index pallet.

• 0xFF01 DXT1  : All Engines except demo
• 0xFF02 DXT2  : Oxygen 2 Only
• 0xFF03 DXT3
• 0xFF04 DXT4  : Oxygen 2 Only
• 0xFF05 DXT5  : Arma1 & 2 Only
• 0x4444 RGBA 4:4:4:4 : OFP and Arma2 only (not Arma1)
• 0x1555 RGBA 5:5:5:1 : Arma1 & 2 Only
• 0x8888 RGBA 8:8:8:8 : Oxygen 2 Only
• 0x8080 GRAY w Alpha : All Engines except demo

'Index Pallets only occur in ofp and demo Later versions of TGAViewer still 'understand' them IF and only IF they have .pac extensions. Under this circumstance they are translated as ARGB8888 (and saved as such)

Note that information on DXT formats can be found at https://www.khronos.org/opengl/wiki/S3_Texture_Compression

Taggs

Taggs do not exist for demo pax files.

Tagg 
{
 char   signature[4]; //always "GGAT"
 char   name[4];      //name of the tagg in reversed order (for example: "CGVA")
 ulong  dataLen;
 byte   data[dataLen];
}

Taggs are read until no "GGAT" signature is encountered. Palette structure follows.

General

Structurally, all TAGG's are optional and can occur multiple times. In practice:

• At least one TAGG exists for ALL types, including index palettes.
• Any TAGG if it occurs, only occurs, once.
• TAGG order of presentation is

AVG  //Always First
MAX  //Always 2nd etc...
FLAG
SWIZ or PROC (not seen together)
OFFS //Always last

• ALL TAGG's begin with an AVG with the sole exception of *some* GREY .paX's
• MAX is almost follows an AVG
• FLAG, PROC, and SWIZ are entirely optional. They may or may not occur.
• PROC if present is only used in DXT1 types
• OFP Palette Indexes ONLY have an AVG
• OFFS is almost always present (if not, the engine builds it as required, when required, anyway)

AVGCTAGG

Average Colour

• This tag contains average color of texture, probably used in rendering the 8:8 Grey & alpha textures.

• Always present, and always first in list with the sole exception of *some* GRAY .paX's

{
  char  "GGATCGVA";
  ulong len; // 4 bytes
  ulong RGBA;//FF443D39
}

MAXCTAGG

• Generally always present. Always follows immediately after an AVG

{
  char  "GGATCXAM";
  ulong len;   // 4 bytes
  ulong  data; // FFFFFFFFF no other value seen so far
}

Contains color of brightest pixel in texture?

FLAGTAGG

• Optional, but will always occur immediately after a MAXC when present

{
  char  "GGATGALF";
  ulong len;   // 4 bytes
  ulong range; // 0 to 2
}

Marks if texture contains transparency. Value 1 means basic transparency, 2 means alpha channel is not interpolated. This flag should be always present in LOD textures with 1-bit alpha with value of 2 or there will be "ghost outlines" on LOD textures when viewed from distance. Note that this flag must be present in texture file when binarizing model, because Binarize stores information about how to render textures in actual P3D file.

SWIZTAGG

• Optional

{
  char  "GGATZIWS";
  ulong len;   // 4 bytes
  ulong  data; // 0x05040203
}

Swizzle is apparently used to modify texture components processing like swizzle modifiers in pixel shaders. For example ArmA sky texture has green channel stored in alpha channel and inversed to take advantage from feature that in DXT5 64 bits are used for alpha channel in each block and 64 bits for RBG, giving double the accuracy to green channel as opposed to storing texture just normally.

Exact format of swizzle data is still unknown. swizzle data:

char channelSwizzleA; 
char channelSwizzleR; 
char channelSwizzleG; 
char channelSwizzleB;

format of swizzle char:

bits 7-4 = 0;
bit 3 - "1" flag. All channel data must be set to 0xff;
bit 2 - "negate flag". Channel data must be negated. 
bit 1-0 - number of color channel:
00 - Alpha channel
01 - Red channel
10 - Green channel
11 - Blue channel

for example (*_nohq.paa textures)

swizzle data is:

0x05 - Alpha, Negated, stored in Red; 
0x04 - Red, Negated, stored in Alpha; 
0x02 - Green, Stored as is; 
0x03 - Blue, Stored as is;

PROCTAGG

• Only ever seen with some DXT1 types

{
  char  "GGATCORP";
  ulong len;   //  strlen(text)
  char  text   // NOT ASCIIz
}

this is a non asciiz string (not zero terminated)

example:

x = ((u+1)*0.5);
y = ((v+1)*0.5);

if (y<0.5) then
{
 sharpOut = 20;
 sharpIn = 60;
 
 offset = x-0.5;
 if (offset<0) then
 {
   edge = (1+offset)^sharpOut;
 }
 else
 {
   edge = (1-offset)^sharpIn;
 };
 res = edge*y*2;
}
else
{
 sharpOut = 20-40*(y-0.5);
 sharpIn = 60-125*(y-0.5);
 
 offset = x-0.5;
 if (offset<0) then
 {
   edge = (1+offset)^sharpOut;
 }
 else
 {
   edge = (1-offset)^sharpIn;
 };
 res = edge;
 sdCoef = y * 2 - 1;
 shoreDisappear = 1 - sdCoef^2;
 res = res * shoreDisappear;
};
a = 1;
r = res;
g = res;
b = res;

OFFSTAGG

• Almost always present

{
  char  "GGATSFFO";
  ulong len;         // 16 * sizeof(ulong)
  ulong offsets[16];
}

Example:
= 6 entries. last 10 unused
256 x 128 Size=16384
128 x 64 Size=4096
64 x 32 Size=1024
32 x 16 Size=256
16 x 8 Size=64
8 x 4 Size=16

MipMap data is presented in 'blocks'. One or more 'blocks' exist in a pax file.

This tag declares where each of these blocks are in the file, relative to start of file.

The location of each block is already known, relative to the size of the previous block (if any). So, although almost always present in pax files, it is use, is redundant.

This tag always contains 16 ULONG offsets. Each one is a hard offset to actual mipmap data relative to start of file.

Not all entries are used (obviously) since most pax files contain less than 16 mipmaps. Unused offsets contain the value 0x00000000. There are no known examples of splattered offset entries. All offsets after the first 0 entry are 0 as well.

Palette

Palette
{
 ushort nPaletteTriplets;                // always 0 except for index palette type
 bytes  BGR_Palette[nPaletteTriplets][3]; // only exists if nPaletteTriplets > 0
}

palette triplets when they exist, consist of BGR values. Note the reversal from the expected order.

Mipmap

Mipmaps are in contiguous blocks that extend to end of file.

There is nothing strange about Mipmaps. Each mipmap is simply a rectangular texture (almost always square) and each one conforms to 2^n dimensions.

Each mipmap defines increasingly poorer resolution. The highest (first) mimpap is THE texture, each successively smaller one is one quarter the size. It progresses in smallness down to a minimum of 2 x any other 2^n or, any other 2^n x 2. For a square texture, this is simply a 2 x 2 matrix.

Mipmap
{
	ushort	width;		// width of this mipmap
	ushort	height;		// height of this mipmap
       if (INDEX PALLETE)
       {
         if (width==0x4D2 && height==0x223D) // special 1234 x 8765 signature
         {
	    ushort	width;		// actual width
	    ushort	height;		// actual height
           // use signed lzss compression for this block (introduced for for resistance pax files.
         }
     //  else use runlength Compression for this block //used by demo and cwc pax files

       }
       if (width && height)
       {
	 byte	size[3];	// size of texture data in file. this is 24-bit unsigned integer. 
	 byte	data[size];	// actual texture data compressed or otherwise
       }
};

The last mipmap is a dummy consisting of zero width and height with NO FURTHER DATA

The size triplet reflects the size of compressed data in the file not necessarily the actual size of the data.

Compression

ALL non DXT signatures are unconditionally compressed using signed LZSS. Index Pallets use this too, OR, runlength encoding.

There are two wrinkles to LZSS compression not encountered in other uses of the same algorithm (pbo, p3d, wrp eg).

• The >1024 rule does not apply. ALL data is compressed. The end result is the smaller box sizes are most often larger on disk.
• The checksum following the data (all data sizes are actually -4 their 'real' length) is SIGNED addititive. All other lzss checksums used elsewhere in bis are UNsigned additive.

example:

Type: RGBA 4:4:4:4 texture
AVGC: F5A9A9AB
256 x  128 Real Size 65536. Size in file  11770
128 x   64 Real Size 16384. Size in file   3521
 64 x   32 Real Size  4096. Size in file   1148
 32 x   16 Real Size  1024. Size in file    407
 16 x    8 Real Size   256. Size in file    169
  8 x    4 Real Size    64. Size in file     57
  4 x    2 Real Size    16. Size in file     22 <<<<<

the result of decompression of any type (lzo, lzss or runlength is encoded data. The encoding is, perhaps obviously,dependent on the paxType:

encoded sizes are a further means of compression.

• width x height is the number of pixels to display. Since all output to a screen is ultimately expressed as RGBA 4 byte integers.
• width x height x 4 is the size of the ultimate output IN BYTES.

Not so the encoded array.

• INDEX width*height // each byte is an index to a color table
• DXT1 width*height/2 // 1 nibble per pixel
• DXT2 width*height //not used
• DXT3 width*height //not used
• DXT4 width*height //not used
• DXT5 width*height // one byte per pixel

• RGB4444 width*height*2 // 4 nibbles per color
• RGB5551 width*height*2 // ditto
• RGB565 width*height*2 // ditto (not used)
• GREY width*height*2 // one byte alpha, one byte 'grey'
• RGB8888 width*height*4 // straight rgba output (un-encoded)

Before ARMA2

DXT textures are stored "as is" in the file. Therefore the DXT data could be directly passed to the graphic hardware. For software decoding of DXT textures see DXTn compress/decompress

POST ARMA2

DXT formats are potentially compressed.

(All other formats remain the same with unconditional lzss compression)

ArmA2 introduces LZO compression for DXT. The criteria for DXT-LZO compression is having the top bit of the width paramater set.

Note that under this circumstance, the width bit must be masked before any size calculations

This bit is set when size on disk for dxt formats != expected size . (See above table). It is either a 'safety bit' for the unusual circumstance of size being same, or, simply a flag to indicate lzo compression.

Either way, there are (currently) no known instances of dxt compression where sizes are same, and quite obviously, all unequal sizes must, by defintion, be lzo compresssed.

Note also, that this is, truly, conditional compression. Only larger dimensions (256x256) eg are treated in this manner. Lesser value mipmaps in the same pax file, remain un altered.

Index Palette Compression

If the file doesn't start with a known PaX type the data array of the mipmaps contain the indices to the color palette (which appears just before mipmaps and only for palette type).

TWO (2) types of compression are in force determined by the special structure of index pallette mipmaps (see above)

• Standard LZSS.

if the signature for width and height IS 1234 x 8765 then standard lzss compression is used, using, the next four bytes as width and height.

otherwise:

A very basic "compression", which even can make the data slightly bigger, is used. The data is split into blocks of the following structure:

struct block
{
 byte flag;
 byte data[...];
}

The flag can be of two different types. If the most significant bit of flag is zero, it tells you how much bytes to read. If flag is 0x05 for example you have to read 6 bytes (always one more then the value of flag). If the most significant bit of flag is one it tells you how often the next byte is repeated. If flag is 0x82 for example the value after the flag byte is repeated 3 times (always one more than (flagvalue - 0x80)).

Note that with this, special, compression, there is no checksum.

Index palette Mipmaps can have mixtures of both types.

Addenda

Decompression Code

see Compressed LZSS File Format

HexDump

The 0x4747 format mentioned in this picture indicates index palette format, because 0x4747 is the "GG" of "GGAT" signature.

Alpha channel interpolation

These two images visualize difference between alpha channel interpolation (FLAGTAGG header tag value).

FLAGTAGG = 1, interpolated alpha channel (default behaviour)

FLAGTAGG = 2, alpha channel interpolation disabled

Bibliography

Feersum's original posting on BIS forums: Paa/pac texture format documentation
MSDN documentation on DXT1 textures: DirectX: Opaque and 1-Bit Alpha Textures
Squish Compression DXTn compress/decompress