Compression is an extremely important effect in sound recording and production. In a typical commercial recording there will probably be compression on every track, and on the mix as a whole. Compressors are often the most invested-in outboard equipment in studios too, with racks filled with valve compressors of varying vintages, and values in the thousands of pounds.
So what does compression actually do?
Compression: the basic principles
Compressors basically reduce the dynamic range of a signal. Volumes over a certain threshold are reduced, removing spikes and peaks in a recording. If you read what I said about the Turn Me Up! Campaign in my post about using a limiter, you maybe wondering why I’d be advocating the use of compression at all…
Well, that problem was concerned with over-compressing an entire song so that it sounds as loud as possible, which leads to less recordings of reduced musicality and dynamic range. However, a certain amount of compression is needed on individual tracks to ‘tidy up’ the performance, and allow the parts to ’sit’ better in the mix.
The best way to describe the correct use of compression is that it tightens up a part, makes it sound more professional and allows it to sit in the mix much more easily. There’s a huge range of compressors on the market; whether a physical box or a digital plug-in for your DAW. Most of them have the same parameters though, which are as follows:
| Threshold | This is the above which the signal is reduced in volume |
| Ratio | This is the ratio of input to output. So a ratio of 10:1 would mean the output signal is reduced by 10 times once it is over the threshold volume. |
| Attack | The time it takes for the compressor to start acting on a signal. This can be instant, or you can choose for the compressor to take a certain amount of time to reach the ratio of compression you want. |
| Release | The time it takes for the compressor to stop acting on a signal once it is below the threshold. Often compressors have an “auto” setting for attack and release, making it easier for beginners. |
| Make-up Gain | Given that the signal is being reduced in volume, often a gain control is added to bring the whole signal up in volume. |
As well as being used to tidy up the dynamic range of parts, heavy compression can be used as an effect in itself, especially on drums. Have a listen to this:
This recording was made using four microphones: two above the drum kit (overheads) and two about a meter away from it pointed towards it (rooms). The room mics have a large amount of compression on them, with a high ratio and low threshold, so everything is blasted to the same volume. The result is a powerful, pumping drum sound.
Thanks for the science lesson, but how do I use it?
Given its subtlety, compression is one of these effects that you have to really fiddle with and explore. Try and take into account the acoustics of the instrument you’re compressing. For example, you might like the attack of the snare, but want to fatten up the rest of the sound. So use a fairly long attack time so the compression is only applied after that first snap of the snare.
It’s probably best to use a fast attack and release on vocals, so the effect isn’t particularly obvious, but does the practical job of reducing the dynamic range.
There is a slightly more interesting use of compression though, and one which you no doubt will have heard in practically every dance song since the early 90s, which is known as…
Side-chaining
Side-chaining involves having a compressor on a track, but using something else as the source of compression. So for example, if you have synth track with some fairly heavy compression on it, but use a kick drum as the side-chain input, the track will only be compressed when the kick drum sounds. You probably won’t be able to do this in Garageband, but it’s certainly possible in Cubase and Logic. Here’s the method:
1. Send the source (kick drum in this case) to a bus.
2. Add a compressor to the track you’re effecting (synth in this case)
3. Set the side-chain input (often called key input) as the bus you sent the kick to
4. Fiddle with the compressor settings until you’re happy – you may want to start with a preset
Here’s a video demonstrating this (you might want to use the ‘full screen’ button on the bottom right of the player to catch the details of what I’m doing.)
And here’s a section of a typical dance remix. Try and spot the side-chaining, there’s quite a lot of it!
I hope this has helped explain some of the mysteries of compression. As I always say, have a good practice and play around yourself, you’ll gradually get used to the effect and it’ll start to sound a bit more noticeable! Next time I’ll start to look into recording techniques for different instruments.
Graeme Allen is the MD of Amati Studios, a company that works with some of the best mid-sized recording studios in the country.
In addition to his technical skills, Graeme is a first class musician in his own right, playing rock, blues and jazz guitar and classical violin. Visit his website at www.amatistudios.co.uk.
