Comment by correa_brian

Yeah, exactly. We capture the semantic meaning of each frame and complement the filename/metadata, so both options work.