CompuServe Messages

That is an issue that I've often thought about. Isn't it possible for a renderer to analyze which elements of the screen move between frames, then set up a bounding box around the elements that do move, and only render them? And speaking of bounding boxes (the mind sure works in mysterious ways…), what do you thing of the "Level of Detail" discussion in chapter 10 of "The Renderman Companion"? Do you think that would be a practical request for r3? Perhaps a LOD setting/switch to be used with either user defined models or interpolated models… You know, I guess the only way your going to satisfy me is to put me on the payroll <GD&R>… -MM