A video would be tough because you'd have to make sure that the camera angle was right. I would say get an image of an eye, put it in Shake and then use a combination of color correction and gamma adjustment to make it "flash". Then slowly add more layers to achieve a more detailed animation such as a slight glow about the retina, maybe a distortion or pattern overlayed on top of the eye, etc. That's the kind of thing that takes a lot of tweaking. It's not difficult, just time-consuming.
When I'm doing stuff like that I put them all into a MultiLayer node because it is very easy to change the blending mode (Add, Subtract, Max, Min, etc) to see what looks best.
Message was edited by: Jon Chappell
****, that's what you get when you use an RSS reader... Sorry I thought this was the Shake forum.