Hmmm tricky. But I think it can be done.
The problem is that you now can't collide your animation bitmaps, rather, a second bitmap that represents the "base area" of your bitmap.
Easy, make a base bitmap for each actor. When the actor is created it spawns its own base actor. The base actor "follows" the animation actor. All collisions are detected by the base actor which then, in turn, modifies the animation actor.
Is there an easier way of doing this? anyone? anyone? bueller?
How does the "parent" thing work. Can an actor refer to and modify its parent?
Should the base animation be the parent and the eyecandy animation actor is the child?
These are the questions that keep me awake at night...
Love to all.