Compression is used extensively in most modern musical styles and is a great tool to have in your skill set. Compression is a complex tool and while it is (fairly) easy to get your head around it, it takes a lot of time and practice to master. Even I still have a long way to go, but I want to share what I have learnt so far anyways :)

What is compression and why should you care?
Rather than just talking about it, let us use a concrete example instead. Listen to the following bass line!

I hope that does sound awful to your ears – because it is! The dynamics of the bass are all over the place, there are really soft and really loud notes played erratically and it’s hard to get a good feel for the rhythm. If you look at the waveform of the bass line, you can clearly see the abrupt changes in volume.

Now if you ever record a track like that, I’d urge you to “flush and re-record”. Working with a horrible recording makes mixing difficult and your final sound quality will suffer if you end up processing the signal too much! However, for the sake of this example, let’s assume you have no chance to re-record and are stuck with this inconsistent bass line. Good news is that we can apply compression to even out the dynamics of the track :)

Essentially, compression takes all sound waves above a certain volume threshold and reduces their excess volume by a specified (compression) ratio. It is important to remember that only the overshoot will be compressed, not the entire signal. This allows us to push down the peaks of a signal and reduce the volume fluctuations of the instrument track.

For our bass line example, I will start off with a threshold of -24dB and a compression ratio of 1:4. This will cause any sound louder than -24dB to have its excess volume reduced to 1/4th of its original output volume.

The red lines mark the -24dB line that we will be using as our compression threshold.

Notice that after compression all levels that exceeded the -24dB threshold have been reduced.

Listen to the compressed version.

You should be able to hear that the overall volume fluctuations of the bass line have been reduced and the bass line sounds more consistent. Notice that the overall volume of the signal has diminished. However, by compressing the peaks of the signal, we have freed up some headroom so we can increase the gain of the entire signal by a certain amount, called the make-up gain, to compensate for this loss in volume. I will use a make-up gain of 8.3dB to bring the level back up to where it was before compression. Be careful to avoid clipping when using make-up gain.

The following image shows the waveform of the compressed bass line after the make-up gain has been applied.

Again, here is the audio example. Compare it to the original bass line.

It’s not perfect, but can you hear how much more even the sound of the bass line has become? Compression is a great way to control the dynamics of an instrument and even them out. The loudness of the overall bass line has also increased because we pushed down the peaks and raised the level of the entire signal up to compensate.

One thing you have to be wary of is overcompression. Too much compression can lead to a very flat and dead sound and, because you typically apply make-up gain to make the quieter parts more prominent, you also increase the volume of any noise in the signal and decrease the signal-to-noise ratio (which is not a good thing). Here is an example of the above bass line overcompressed with a threshold of -40dB, a compression ratio of 1:8 and a make-up gain of 23.5dB.

Notice how our threshold is so low that almost the entire signal will be compressed and not just the peaks.

The waveform of the overcompressed bass line looks almost like a square – all dynamics have been squeezed out of it.

Have a listen to this overcompressed bass.

In this example, the life as well as the natural sound of the bass has been squeezed out of the bass line and a lot of low noise that was present in the signal has been increased in volume by using too much make-up gain.

While compression has a distinct sound that experienced sound engineers can pick up on very easily, most people (including me to some extend) will hardly notice even strong compression. I do hope you notice the ugly compression sound in the overcompressed example tho ;)
Compressed instruments usually sound thicker, more compact and punchier, which makes compressors popular to use for kick, snare or bass tracks. It is also used for vocal tracks, but simply to even out the dynamics a little like we did in this example.

In order to preserve the natural sound of your instrument track, it is usually better to compress a signal more frequently by a little bit than to compress it a lot once.
I like to start off with a fairly high threshold (so little of my signal is affected) and a small compression ratio (approximately 1:4 is fairly common) and then adjust your parameters based on what it is I am mixing. Compression is best learnt through experimentation and feel free to download the above bass line and play around with it :)

Wait!
Stop!
You didn’t mention the attack and release parameters!

Yeah I know, but I’m feeling like this post is already getting way too long and those compressor parameters are important enough to warrant a separate post to cover them properly. I will write about them soon – promise :)