The hardest part in any audio situation is getting the compression right. Every source has a different need when it comes to compression, and it is almost impossible  to give a universal setting for a given audio source, especially when everyone approaches his/her instrument (including voice) in a different way.
However, here is a little information on setting up compression that can be helpful. Start by running the raw source material through the compressor and have the process set to "out" so that the signal is untouched dynamically. Then, set the ratio to about 5:1 or 6:1, and the threshold all the way to the right. Set the process to "in" so that the processing can be heard. Start decreasing the threshold until you see around 3-6dB of gain reduction. If that's not enough compression, then lower that threshold again until -6 to -10 dB of gain reduction is noticable. At this point, it should sound a bit overly compressed, and when switching from "in" to "out" it should be possible to hear a big difference. Then from this point, start to dial in a slightly lower ration, and a slightly higher threshold, until the compression is smooth and natural sounding (continue to use the "in"/"out" as a reference).
Compressors are "input level dependent" devices - e.g. changing the input signal level even slightly will change the effect of the compressor, even when the physical settings of the compressor's parameters remain unchanged. So, once the input level is set, don't change it. Instead, change the output level on the compressor to "make-up" the level.