I'm aware of motion blurring via the accumulation buffer, although relatively simplistic (afaik from red book), it will be rather slow since it isn't support in hardware. I take it teh best way is to take a snapshot of the previousscene and then reapply this to current scene
Could this be applied to say streaming videos
i.e. take snapshot of the video, then apply this as a 2D texture to a quad, so you can still do roations etc, but the video will still play?
Hi res videos will be a LOT slower but low res videos maybe would work???