Compression explained with diagrams

Chances are you've heard of compression. But what is it and why use it? This article will answer both those questions.

Compression is a way of balancing the dynamic range of a signal. To put that a simpler way, compression is used to make an audio signal sound more even overall. Compression can be applied to anything - from a piano to a drumkit to a vocalist. Or even a chirping bird or roaring race car engine!

One of the main reasons compression is used is to stop signal peaks hitting the 'red line', that is becoming distorted because of signal overload. You can see that this signal has been overloaded and 'peaked'. This is also called 'clipping'.

distorted audio signal meter

One solution would be to turn down the signal. The problem with that is that the whole signal gets turned down, including the quiet bits. Using a compressor can solve this problem. Another popular reason for using it is to balance an entire signal. When you record a guitarist or a vocalist some parts of their performance are going to be louder than others, and some bits quieter that others. That's called 'dynamic range'. Reducing dynamic range often makes a far better signal.

Compression can be applied to live performances or something pre-recorded. Simply plug a compressor into either situation and you're all good. Except, you need to know how to use one right?

Take a listen to this brief audio sample. It's an acoustic guitar. A full C-Major chord is strummed, then the strings silenced, then some notes picked. It could be a part of any song. The strummed chord is much louder than the picked notes.

Audio signals are audio waves. Almost always there is an initial hit and then a trailing off of the sound. That trailing off part is called a 'transient'.


Here's the thing. All the good stuff is in the transient. Another part of audio signals is peaks. Those are the loud bits of a signal. In our acoustic guitar sample they occur each time a note or chord is played.

Audio signal with peaks

The problem is if we boost the audio signal to make the transient louder we'll also be boosting the peaks. That's going to make the audio signal worse! What if we could turn up the transient and turn down the peaks at the same time?

high and low parts of audio signal

The solution is to use a compressor. A compressor does exactly that. It reduces the peaks and boosts the overall signal. There are a few parameters to using a compressor. This article will deal with only two - Threshold and Ratio. You'll get by quite nicely if you can understand these two parameters. They work the same whether you are using a piece of rack-mounted hardware or a digital version on your computer or device. Once you get the hang of one compressor you'll be able to use any without much hassle.

As we mentioned, one parameter is Threshold. This indicates at which point of the signal we want the Threshold to engage. At which part of the 'loudness' of the signal we want to start reducing it.

compresssor threshold diagram

You can set Threshold anywhere you like for processing your audio signal. For example, you could have a loose (or, high) Threshold. If you set your threshold high like is shown in the next diagram it would only engage on those tiny little peaks outside the threshold.

high threshold on compressor

If you set a tight (or low) Threshold your compressor would engage very quickly. Compression can be used both creatively and logically. Generally, you only want just as much compression is needed. Some engineers, artists and music genres use heavy compression. One reason is to help each instrument sit in a good place in the mix. Let's focus on compression by using a mash-up diagram. We've established a threshold. We want to reduce those peaks and boost the transient.

peaks and transient in audio signal

It should be noted that when you are using a compressor you will not see your audio signals being presented like they are in these diagrams. The diagrams are to help you understand how a compressor works. Often, compressors will indicate they are working in some visual way, perhaps a light that engages when the threshold is reached, or perhaps a floating icon.

Once you have worked where you want your compressor to engage (the Threshold) you then need to tell it how much to adjust the signal. This is called the Ratio. Everything inside the Threshold will be unaffected. Everything outside the Threshold will be affected - in our diagrams those would be the peaks.

How much do you want the compressor to reduce those peaks? A little? A lot? A low ratio (2:1, 3:1, etc) affects the signal outside the threshold only a little. A high ratio (6:1, 10:1, 12:1, etc) affects the signal outside the threshold quite a bit. 20:1 (pronounced 'twenty to one') is called 'limiting compression'. Your audio signal is extremely reduced. Generally, the average person won't use such extreme compression but it can be useful to ensure that a signal never exceeds a certain threshold, perhaps for broadcast media or live performances.

The next diagram shows our original acoustic guitar signal given a mild compression. The threshold is reasonably average, and the ratio is about 3:1. Comparing the diagrams you can probably see where the Threshold was set. You can see everything inside the threshold is mostly the same (remember, that's all the good stuff) and the peaks have been reduced, but not destroyed.

diagram of audio compression in music

Take a listen to our audio track which matches the diagrams. First you hear the original acoustic guitar signal then you hear the one that has been compressed

How much compression to use? Well, that's the question isn't it. Like we mentioned, some people like it light, some heavy and the situation can also determine how much compression to use. You might think that the original acoustic signal in our example sounds nice. But, remember that first peak is quite intense. It will interfere with the overall level so it needs to be compressed. The compressed version is much better. It retains almost all of the acoustic qualities yet is a nicely balanced signal.

Don't be afraid to experiment with your compression. Often you will find that just enough compression is exactly what you need.

All of the screen shots so far have been from Audacity. We chose to use that because it is a free software that many people use. Others might use a professional platform like Pro Tools. The next diagram shows the Pro Tools compressor window. Many Digital Audio Workstations and even video softwares have a similar interface. We've buffed the graphics a little to help you see different engagements of the Threshold and Ratio. You may need to zoom in to see what's going on in all the examples.

Pro Tools compressor images

To sum up - Threshold is where the compressor begins working (think 'at which point of signal loudness do I want the compressor to start working'). Bascially, a compressor will not do anything until the Threshold is reached. One the threshold is reached how much do I want the compressor to reduce everything that is too loud, the Ratio.

Lastly, our audio sample only went for a few seconds. You will probably be using your compressor to affect your bass guitar for an entire live performance, or smooth out the full vocal track of your digitally created song. Generally, a compressor is set-and-forget. Once you've found the settings you want leave it alone and it'll do it's job quite happily.

So that's everything. I hope you found this article useful. If you did please give it a like and even  a comment.

Happy Jamming.

IK Multimedia's T-RackS 5