The compressor is one of the most useful tools a mix engineers has at their disposal yet it is also often one of the least well understood. It doesn’t help that the term compression is also used to describe the process of creating an MP3 from a WAV file, but what we’re talking about here is dynamic compression — a process that helps us keep control of varying signal levels.
Before compressors were invented, the only way to control the level of a signal was by hand — if it was too quiet you’d push up the fader and if it was too loud you’d pull the fader down. Viewed at its most simplistic, this is pretty much what a compressor does for you, though the physical fader is replaced by some form of electronic gain control circuit and there’s more electronics to monitor the incoming signal level and to adjust the amount of gain control. Of course a correctly designed electronic circuit or software algorithm can react to changing levels much more quickly than a human engineer, but some human intervention is still required to set up the compressor in the first place.
Though there are several variations on the basic compressor theme, many covered in detail in back issues of Sound On Sound via the soundonsound.com website, the most common system employs what is known as a threshold, a level below which no action takes place but above which the signal gain is reduced. This is the equivalent of our engineer moving the fader down only when he thinks the level has got too loud. A typical compressor will have a rotary control for threshold calibrated in dBs with a gain reduction meter of some kind to show when and by how much the signal is being compressed.
The other user adjustable parameters found on a typical compressor comprise Ratio, Attack, Release and Make-up gain, though not all models have all of these controls and some may have additional controls not described here. Ratio determines by how much the gain will be reduced once the signal exceeds the threshold. With a ratio of 1:1 there is no compression at all while a 4:1 ratio means that the input signal level will have to rise 4dB above the threshold level for the output signal to increase by only 1dB. Essentially the ratio value tells you how many dBs of input level increase is needed to result in a 1dB increase in the output level for signals that pass the threshold level. If you set a very high ratio, the signal level is effectively prevented from exceeding the threshold level at all (assuming that the compressor can work extremely quickly), and that’s exactly what a Limiter is — a very fast compressor with a very high ratio.
Attack and Release adjust the speed with which the compressor reacts once a signal passes through the threshold. A short attack means that gain reduction is applied very quickly while a slow attack allows the signal to exceed the threshold value for a few milliseconds before the gain reduction circuitry reins it in. This can be useful for preserving the attack of percussive or plucked sounds.
Release determines the time it takes for the compressor to stop applying gain reduction once the signal has dropped back below the threshold. This prevents the level changes being too abrupt, which might otherwise result in an effect known as gain pumping where you actually hear the compressor changing the signal gain. Deliberate gain pumping is sometimes used to add an element of excitement to rock and dance mixes so give it a try by setting short attack and release times, a high ratio and a threshold low enough to to give you 12dB or more of gain reduction on the loudest peaks.
Because a compressor applies gain reduction to any signal peaks exceeding the threshold level, most designs include an adjustable output gain or make-up gain to bring the signal level up to its previous peak value. The outcome of bringing up the gain in this way is that the peaks are brought up to the same level as before while quieter sounds are now made louder so there’s less level difference between the loudest and quietest parts of a track.
Note that many plug-in compressors come with a list of presets suited to different types of sound — vocals, percussion, guitar and so on. These can be very useful with one proviso — you will still need to adjust the threshold control manually. That’s because the person who designed the preset doesn’t know how loud your signal was recorded or how much compression it needs. The best way to adjust the threshold is simply to watch the gain reduction meter with a loud part of the track playing, then adjust the threshold so that the meter is showing around 5dB of gain reduction on the loudest peaks. You can then move either way from this value according to the style of music. More gain reduction means more compression and a greater reduction in dynamic range, but if you go too far the process of compression might drain some of the life out of your sound rather than making it bigger. Only your ears can tell.
Compression in the Mix
When mixing, you might use a compressor to smooth out level differences in a track, but you can also use the Attack and Release controls to modify the envelope of a percussive sound. As mentioned earlier, a longer attack time allows more of the attack of the original sound to get through unprocessed, which may produce more definition to the sound, while a fast attack tends to bring transients under control almost immediately resulting in a less pronounced attack to the sound. A slower attack can be useful lending more definition to a drum or picked bass guitar.
When processing repetitive sounds such as drums, it is important to set the Release time fast enough so that the gain reduction returns to zero between hits.
Some models of compressor include an ‘Auto’ setting that continually adjusts the release time automatically depending on the dynamics of the incoming signal. Auto can be a good choice when you want the compression to control your dynamic range without sounding too obvious. Auto can also be useful when processing complex mixes or instruments where the attack and release characteristics aren’t consistent and where a single fixed release setting might not be the best solution.
A compressor evaluates the level of the incoming signal in a piece of circuitry known as the side-chain, but what’s the best way to work out the level? The human hearing system tends to average out sound levels so that short duration loud sounds don’t appear to be as loud as longer sounds of the same peak level. The RMS measurement system used by many compressors treats levels in much the same way as the human ear by tending to average out brief peak levels, and this is ideal for getting non-percussive material to sound more consistent in level. However, it isn’t necessarily the best approach for processing drums and other brief percussive sounds, which is why some compressors allow you to switch from RMS to Peak level monitoring. In Peak mode the compressor will react to high level sounds no matter how short they are. As a very general rule, Peak is most effective on short-duration transient sounds while RMS is best suited to vocals and sustaining sounds.
There’ll be more compression tips from me next time, so check back in a week to learn some of the finer techniques, including compression on vocals. See you then.