Have you used the parallel compression on the SQ to limit the DR of people talking.
EG speaking in a church or lecture in an auditorium. Did you use it for people singing too?
How much did/could you squash the DR and what settings did you use when you accomplished that?