Parallel compression is often talked about in hushed tones, as if it carries almost mythical powers to improve your mix. However, in reality it is relatively straightforward. At its simplest it involves mixing compressed audio with an unprocessed version of the same signal (hence the name).
This can be achieved in a number of ways, with the most convenient method partly determined by your choice of DAW or compressor plugin. For example, if you’re dealing with a single audio recording of all the drums, it’s simply a matter of duplicating it on another track and then applying the processing to the new copy. Alternatively, you may also use auxiliary sends pointing to a dedicated compression bus. The easiest way of all is becoming increasingly common: some plugins feature a mix control, making experimentation straightforward.
Although you can try the processing with any settings, a common approach is to set your compressor to dig in quickly with a fast attack and medium-fast release, then bring the threshold down into ‘way too much’ territory. Mixing the original back in will restore the squashed transients, but also pick up the level between hits. In fact, parallel compression often means that you notice the effect most on quieter elements, as the main hits are now much louder than their squashed counterparts.
We spoke in greater detail about parallel compression in this walkthrough.