Some of the items used every day in the recording studio make sense on an intuitive level: The primary function of a microphone is to convert sound waves to electronic signals; the primary function of a preamp is to make quiet electronic signals louder; and the primary function of an equalizer is to change the sound of a signal by boosting or cutting specific frequencies. Of course, there’s much more to it than that, but these generalizations get the point across as a starting point. And then, there are compressors and limiters, devices whose function is often described in almost lyrical terms: “bottom,” “fatness,” and “edge” are some of the more poetic (and printable) descriptions of what someone’s favorite compressor does. But since those terms don’t help to explain what a compressor does, here’s an explanation.
A compressor’s primary function is to reduce dynamic range — that is, the difference between the loudest and softest parts of a track. This is typically done by clamping down on the loudest parts of a track (gain reduction) then bringing up the overall level of the track (makeup gain). Though some compressors have controls for only these two actions, it’s become common for manufacturers to offer a bit more control over other variables. The following is a discussion of what those variables are and why they matter.
Compressors allow audio signals to pass unaltered until the level reaches the threshold, the point at which the compressor circuitry starts reducing gain. Below the threshold, there’s no change; above the threshold, the signal starts being attenuated. Some compressors have a preset threshold, while others allow the user to choose. If the threshold is set to -12dB, then any time the signal rises above -12dB, the compressor starts to reduce gain by a set amount (usually written as a ratio).
The ratio is used to describe the amount of gain reduction that is actually occurring. If a compressor is set to a 2:1 ratio, the signal increases by 1dB for each 2dB of gain above the threshold. So, if a signal peak coming into the compressor exceeds the threshold by 4dB, the output of that peak will only be 2dB louder than it would be if unprocessed; the gain is reduced by 2dB. If the signal exceeds the threshold by 8dB, then the compressed signal will be only 4dB louder. At greater ratios, the amount of gain reduction is correspondingly different: With a 4:1 ratio, the output signal increases 1dB for each 4dB of signal strength over the threshold. In general, a 2:1 compression ratio is quite mild. A 4:1 ratio is moderate, and an 8:1 ratio is pretty heavy. Ratios of 10:1 and greater are considered limiting. This means that the difference between a compressor and a limiter is the ratio. A limiter is a compressor that operates at ratios greater than 10:1, so every limiter is a compressor, but not every compressor can be a limiter — only those whose ratios can be set to ratios of 10:1 or greater. Make sense?
The attack time is the time it takes for the compressor to reach the preset amount of gain reduction once the threshold is exceeded. If the compressor has an attack-time setting of 20 milliseconds (a fairly slow attack time), then once the signal is higher than the threshold, it will take 20 milliseconds for the compressor to reach the full amount of gain reduction. The compression circuitry begins working as soon as the threshold is exceeded, but the attack time is kind of like the accelerator on a car; even if you stomp on the accelerator (exceed the threshold), it takes a bit of time to get up to speed (reach the desired amount of gain reduction).
The release time is the amount of time it takes the compressor circuit to stop attenuating the signal after it drops below the threshold. With the release time set at 50 milliseconds, it will take 50 milliseconds after the signal drops below the threshold for the compressor to reach 0dB of gain reduction. Continuing with our accelerator metaphor, if you’re cruising at 70 mph, taking your foot off the gas won’t immediately bring the car to a halt; it’s going to take a little while to slow down.
When the loudest parts of the signal are compressed, the result is that the signal is quieter. Makeup gain is used to bring the overall output level of the compressor back up. So, if your compressor is, on average, reducing the peak levels by 5dB, then you can boost the makeup-gain control by 5dB to bring the output volume back up.
You can buy compressors with tube signal paths or solid-state signal paths, and there is a number of different circuit topologies. Choose between optical, VCA, FET, or tube (Vari-mu) gain-reduction circuits. Each of these designs has strengths and weaknesses, and the different designs can have quite different sonic footprints. But if you start by remembering that the basic function of a compressor is to reduce the dynamic range of a signal — then everything else becomes a little clearer.