You can think of a piece of audio, visually, as a mountain range. It has tall bits, shorter bits and medium bits. You could say it has a DYANAMIC RANGE. Clever.
If I’m standing on the highest peak and try talking to someone on the lowest peak, well, the difference in volume would become apparent to me.
Enter the compressor.
The job of the compressor is to squish down the loudest parts of the audio, so they move closer to the quieter parts, giving a more even overall volume… reducing the dynamic (mountain) range.
If you’re recording your voice and you blurt out some words or parts of words more than others, whack a compressor on it.
Just in case my metaphor didn't make much sense, let's have a quick look at an actual audio file. (I'll never get used to saying "look" at an "audio" file).
This first image is a few sentences of a script. You can see 5 or 6 peaks that stand above the rest. If you listen to the audio like this, those parts will jump out at you. They'll be brief but but noticeable. The job of the compressor is get those spikes looking (SOUNDING) like the rest.
Have a look at the next image which is after the compressor has been applied. Those tall peaks have been reigned in and the overall audio level has been levelled out more evenly.
There's more to do after compressing though, which you can find out about in another post - How to improve your voice recording quality. Scroll down to point 4.