Hello guys. I'm having an issue with using compressors for voice recordings, namely when two voices vary wildly. For example, If a speaker's voice peaks at -20dB and another one's at -1 dB, no matter what compressor settings I use, in the end there still is a big difference in the perceived loudness of the voices.
The only solution that I found so far is to manually raise the volume only on the soft parts. However, this is time consuming and I'm looking for an automated process that can take care of this. Can you guys think of one?