Investigate if there is a way to take advantage of
hardware acceleration (i.e. either hand tuned assembly
to replace generic conversion routines on specific
platforms, some sort of tricksy smartness involving
using the video card to do the conversions for us, etc.)