Controlling pit orchestra monitor levels for musical theater
I recently worked a high school production of a musical. I am a professional musician and an "amateur sound guy". I was the sound designer for the show and played drums in the pit orchestra (a real pit, under the stage). Other folks were the sound operators. We used an A&H SQ7, and the pit orchestra monitor feed was a dedicated aux output on the board, fed by post-fader signals from 24 wireless microphones, a few wired microphones, and the digital audio feed to the board used for the show's sound effects.
Throughout tech rehearsals we were struggling with monitor levels in the pit. I had a unique capability as the show's sound designer to have lots of control over this output via my phone even though I was in the pit. I messed with the compression settings a lot trying to get monitor output that was loud enough to hear during music (actors singing while the orchestra is playing) and not uncomfortably loud during no-music speaking moments. I think I did a good job of dialing in the compression so that quiet singing/playing and loud singing/playing worked without further adjustments, but I was not able to figure anything out to deal with either uncomfortably loud speaking-only, or uselessly-quiet (i.e. effectively inaudible) output during music. Since the latter is absolutely unacceptable, the orchestra was getting blasted by dialogue when we weren't playing.
I came to the conclusion that compression wasn't going to help. This wasn't about actors needing to have more or less sound system output based on their microphone input; this was about needing more or less volume from the monitor loudspeakers depending on the scenario, for the same actor microphone input. So I worked it out with the sound operators that they would give the orchestra 12 dB more pit monitor level during music, and 12 dB less during speaking-only. I kept the SQ MixPad app up on my phone throughout rehearsals and performances so that I could keep an eye on this, tweak as necessary during some quieter moments in the music where we didn't need all 12 dB, and make the adjustment if the operators forgot (it's hard to mix what you can't hear, and this was not a process that we had automated in any way). Sometimes I had to turn the monitor output level up on a tiny phone screen with one hand while playing drumset one-handed, which is obviously less than ideal.
How have others dealt with this situation?