1: Don’t. I would rather spend the time to ride the solo or vocal to get a cleaner sound with no compression artifacts. I also prefer to manually remove pops and sibilance. You can use the volume automation in a DAW to eliminate vocal pops and sibilance problems by drawing a V-shaped notch at the center of the pop or ess. It does not have to be very wide, and it will work better than any automated de-esser or pop filter.
2: For the most transparent compression, use a ratio between 2:1 and 3:1. This will increase the apparent loudness of your vocal, but will not have that annoying pumping sound of badly adjusted compressor settings.
3: Don’t compress more than 4dB. Watch the gain reduction meter on the compressor. Adjust the input gain or threshold level until the reduction reads between 3 and 4dB, no more.
4: Use multiple compressors connected in a series if you need more than 4dB of compression. Set the attack and release settings differently and you will have more compression without sounding like you’re killing the vocalist.
5: Parallel compression works in some circumstances. You have the dry signal and the compressed signal – mix them together to get the sound you want. Make sure you compensate for any delay in the compressor to avoid phasing.
Roger Nichols is a recoding engineer and producer and has won seven Grammy Awards, the 2001 TEC Award, and received 11 Grammy nominations. He is on the Board of Governors for the Miami Chapter of NARAS and lectures at Berklee School of Music, Musicians Institute, Recording Workshop, Full Sail, Vancouver Film School, and University of Miami. Visit www.rogernichols.com.