Should we consider a common spectral framework and would that improve spectral fitting?