contact: phil@flack99.freeserve.co.uk
The Header:
| Bytes | Values | Notes |
| 0-3
4-7 8-11 |
"PAMB"
0x00000000 data size |
Section Type (BMAP)
0 size of following data |
After the header section comes the real header section:
| Bytes | Values | Notes |
| 0-3
4-7 8-11 12 13-16 17-20 21-24 25 26 27 |
"DHMB"
0x00000000 data size image type actual width actual height bytes per row unknown padding padding |
Section Type (Bitmap Header)
0 size of following data Specifies type of BMAP (see later) width of image in pixels (actual, see later) height of image in pixels number of bytes per row (often unexpected size, see later) can't find a reason for this - only seen as 0 padding 0x20 byte padding 0x20 byte |
If the image type is 2, a CMAP follows.
Then the DATA section follows (||ATAD||0000||SIZE||data...)
where data... is the pixel values.
BMAP Type
There are 6 different BMAP types:
| Type | Pixel format | Other |
| 0
1 2 3 4 5 |
4bit, external CMAP
4bit, external CMAP 4bit, internal CMAP 16bit, rgb 565 16bit, prgb 1555 16bit, argb 4444 |
CMAP from preceeding section is used: All Opaque
": One Transparent colour Following the BMHD is a CMAP: Varying transparencies rgb format 5r 6g 5b rgb format 5r 5g 5b (1st bit (msb) can be ignored - maybe transparency) argb format 4a 4r 4g 4b (a=alpha=translucency) |
Bytes per Row
This value gives the number of bytes that must be read *height to read
the image in from file correctly. It is not necessarily the same
as width*bpp/8 (i.e. width*2 for 16bit pixels). If the image type
is 4bit, then each byte contains two pixels (first pixel is byte/16, second
pixel is byte%16). To display each line, read in bytesperrow pixels
into a buffer; working left to right display each half-pixel in turn, ensuring
that the position the pixel is drawn is <= imagewidth. This ensures
that the trailing pixels are ignored. The same goes for 16bit pixels,
but this time two bytes need to be read for each pixel.