I wonder if you could *model* the way the eye *would* track something. That way you don't need a human volunteer every single time you do a transformation. -Keith