Comments on: Omni-directional shadow mapping