chughes@maths.tcd.ie (Conrad Hughes) (03/26/91)
I'm trying to compress programs involving large amounts of
(handcoded, if that makes a difference) machine code, graphics and
sound. I have a distinct feeling that using the same compression
algorithm for all three isn't all that good an idea.. So, anybody
know of any particularly good algorithms for compressing data of the
following form..
32-bit word assembler (each instruction is as usual divided into
several bitfields, some of which (such as the condition bits) have
the same position and width in every instruction. The processor is
the Acorn ARM2 RISC processor).
Algorithm already used: Huffman (or close relative)
8-bit two's complement signed centre-zero sound samples.
Algorithm already used: subtract each byte from the next, and then
apply the above Huffman variant on an 8-bit word level.
If anybody knows of particularly effective algorithms for 8-bit
graphics (either for graphics with a high random element, i.e.
repeat packing won't work, or with reasonably large connected areas
of various shades), let me know also..
If I recieve many email replies I'll post a summary of anything
useful I manage to implement..
Thanks,
Conrad Hughes
--
Will you sell me one of those if I shave my head
Get me out of town is what Fireball said
Never trust a man in a blue trench coat
Never drive a car when you're dead Tom Waits