Hence the name, Run-Length-Encoding.
Sprites have an awful lot of pixels around them which are both black and transparent - this would make a pretty noticeable difference in size if each one was defined independently, right? Why bother making an 'RLE' if it actually isn't?
(EDIT) Tests...
It's true MC RLEs are not RLEs at all. Luckily both MC RLEs and EnRLE RLEs can be zipped down to the same size :). There's a question - how come RLEs (even EnRLE ones) can be zipped down so much smaller?
This post has been edited by Guy : 19 January 2006 - 03:20 AM