From inSync reader Mal M. comes the following question:
"I have a radio interview and wish to process it in my digital audio editor to bring up the level of the interviewee, who was too quiet. How do I set the threshold on the compressor to leave the interviewer alone and bring up the level of the interviewer?"
Tough one Mal! It will be difficult to set up a compression threshold to only process the softer signal; compressors tend to work the other way around, processing signals loud enough to cross the threshold. Off the top of our heads, we might try setting up the dynamics processing so that the louder part is being heavily compressed or limited, then turn up the level of the entire track until the softer signal is as loud as you would like. The danger here is that the louder signal will become so squashed that the processing will be audible (depends on how far apart the two levels are). Use your ears when trying this approach.
You might also play with setting up a sidechain (if your compressor has one) that turns down the louder signal, while letting the softer one pass, similar to the way that compressors are used to control music bed volume when performing "ducking" on voice-overs (see inSync WFTD 5/28/97 for more info on using sidechains as duckers)
Finally, if you are working on a computer-based digital editor with DSP capabilities, you might try selecting the regions that are too quiet and normalizing them to raise the volume while maintaining relative dynamics (On a quality DSP system, this might be the path of least audio quality compromise).