That makes sense, I'd still like to know what kind of footprint the combined decoder needs.