I like using parallel compression and we will come to an example of this later at the mix stage.
The process is extremely simple and entails using two channels, one with a compressor inserted in the channel and set to completely ‘wet’ (affected) and the other running the dry unaffected version of the same piece of audio that needs affecting. In essence, you make a copy of the audio you want to treat and use a compressor on the copied version and fully wet, and then mix this affected (wet) version with the untreated and dry original.
The best technique is to use the bus out or aux on the channel to send the audio to the compressor and feed it’s output back into another channel. However, using the ‘copy’ technique and affecting the copy on a separate channel and mixing to taste also works. Please do not confuse this with side chaining.
The beauty of using parallel compression over a simple a compressor on an insert in a channel is that you have far more dynamic control over the ‘wet’ and ‘dry’ signals. The dry channel maintains the dynamics of the unprocessed audio, and the compressed channel offers colour and body.
Mixing the two affords endless variations of dynamic shaping.
Nowadays, we have compressor with a dry/wet mix which acts as a parallel configuration. A good example of this is Cytomic's Glue compressor whcih I use regularly.