Are you sure your slowdown is not due to the ioctl() interface for userland screen updates? To test the theory, use the regular empeg MMAP function from empeg_display() to run your app while the player is shut down.. see if it's any faster that way (direct memory access, almost..).

That's good thinking. I will give that a shot to see what's up. The thing is, then I can't test it while the player is running, which is when I see more slowdown. There's still some level of slowdown when I run "sans player" but it's not as noticable. I'll see if I can figure out who the culprit is. Thanks for the testing strategy.