Welcome to the Creatures Wiki! Log in and join the community.

Difference between revisions of "BLK files"

From Creatures Wiki
Jump to navigation Jump to search
(More detailed BLK docs)
(fix stray <)
Line 40: Line 40:
 
The following macros can be used in C or C++ to decode these images to normal 32-bit format:
 
The following macros can be used in C or C++ to decode these images to normal 32-bit format:
  
<pre><
+
<pre>
 
#define RGB(r, g, b) ((r) &lt;&lt; 16 | (g) &lt;&lt; 8 | (b))
 
#define RGB(r, g, b) ((r) &lt;&lt; 16 | (g) &lt;&lt; 8 | (b))
  

Revision as of 04:55, 11 February 2005

BLK files are basically S16 files with a slightly different header, and with slightly incorrect offsets due to the different header. They contain 128x128 sprites which are the 'blocks' which make up the background.

The header is as follows (all values are little-endian):

flags32-bit integeronly the first bit is used; if set, the image is in 555 format, if not, the image is in 565 format
backgroundwidth16-bit integerthe width of the background, in blocks (not pixels!)
backgroundheight16-bit integerthe height of the background, in blocks (not pixels!)
numframes16-bit integerthe number of sprites in the file (must always equal backgroundwidth * backgroundheight)

Then, for each sprite:

offset32-bit integeroffset of the sprite data from the start of the file, minus 4.
width16-bit integerwidth of the sprite. must be 128.
height16-bit integerheight of the sprite. must be 128.

The rest of the file consists of image data at the file offsets defined in the header. Images are stored as an array of horizontal scanlines. Each pixel is represented as a 16-bit little-endian integer. The breakdown of the integer into red, green, and blue components depends on the format (555 or 565).

In 555 format, the bits are allocated as follows:

 0123 4567 0123 4567
 -rrr rr-- ---- ----
 ---- --gg ggg- ----
 ---- ---- ---b bbbb

In 565 format, the bits are allocated as follows:

 0123 4567 0123 4567
 rrrr r--- ---- ----
 ---- -ggg ggg- ----
 ---- ---- ---b bbbb

The following macros can be used in C or C++ to decode these images to normal 32-bit format:

#define RGB(r, g, b) ((r) << 16 | (g) << 8 | (b))

#define DECODE_555(v)           \
	(qRgb(                      \
		  ((uint32_t)(v) & 0x7800) >> 7,  \
		  ((uint32_t)(v) & 0x03e0) >> 2,  \
		  ((uint32_t)(v) & 0x001f) << 3  \
		 ))

#define DECODE_565(v)           \
	(qRgb(                      \
		  ((uint32_t)(v) & 0xf800) >> 8 , \
		  ((uint32_t)(v) & 0x07e0) >> 3 , \
		  ((uint32_t)(v) & 0x001f) << 3   \
		 ))