Today I'm going to talk about compressors. What are the main parameters? When do we use a compressor? And as usual, I'm going to give you some practical advice.
What is a compressor?
A compressor is, first of all, a tool that helps us to control and shape the dynamic range of an audio source. In some cases, it can be used to add texture or to glue up a group of instruments.
Compressors can reduce the dynamic of a sound in two different ways:
Downward compression: these kinds of compressors reduce the sound that goes above the threshold. The part of the sound below the threshold won't be affected by the compressor. Most of the compressors work in a downward way.
Upward compression: these kinds of compressors instead increase the loudness of the sound below a certain threshold. The part of the sound above the threshold won't be affected by the compressor.
Type of compressors
Without being too pedant, there are four main types of analogue compressors which involve different technologies. (Plug-ins simply emulate these technologies)
VCA (Voltage Controlled Amplifier): based on VCA circuit
Optical (or Opto comp.): based on light-dependent resistors
FET (Field Effect Transistor): based on transistors
Delta Mu or Vari-Mu: based on valves
PWM: based on Pulse Width Modulator circuits
Depending on the technology involved and the manufacturer a compressor can feature different parameters. We can tweak some of them while some others are fixed.
the level at which the compression starts working. Of course the lower the threshold, the more the compressor will compress the audio (downward compression), but bear in mind that the main purpose of the threshold is to establish when you want the compressor to process the sound.
Some compressors, like the 1176, have a fixed threshold. To compress more you have to rise the gain input to push the signal against the threshold.
Gain Reduction Meter:
how much the gain has been reduced by the compression.
determines how much the volume has reduced when the sound hits the threshold. In other words how much we are compressing. The ratio is expressed in fractions (E.G. 2:1, 3:1, 4:1 and so on).
Let's see some examples:
If I set the ratio to 2:1 means that only half of the signal that goes above the threshold passes through while the other half is attenuated by the compressor.
A ratio of 4:1 means that 1/4th of the sound that goes beyond the threshold passes through. This means that if 4db passes the threshold as a result our sound will be attenuated by 3 db.
A ratio of 10:1 means that only 1/10th of the sound that goes above the compression passes through and so on. When we compress a signal we don't really care about the calculation we are more focused on how the final result sounds. The theory makes us understand that the hight the ratio the more aggressive the compression.
expressed in milliseconds. It determines how quickly the compression reacts when a sound hits the threshold.
You can also look at it as a fade from the uncompressed to the compressed signal.
is the time that the compressor takes to recover. In other words, it is the time that the compression takes to return the audio to its original level. You can also look at it as a fade from the compressed to the uncompressed signal.
determines how the ratio behaves as a signal approach the threshold. With a hard knee, the radio applied is the one we choose as the signal hit the threshold. With a soft knee, the radio gradually increases as the signal approach the threshold. Usually, a hard knee makes the compressor sounds more obvious and a soft knee makes it sounds more natural.
The make-up gain is a simple gain that allows us to compensate for that loss of volume due to the compression.
Detection type - Peak / RMS:
it makes the compressor react to the peaks of the signal or the RMS (average level).
it can be found only in virtual emulations. This feature allows the compressor to look at what is coming ahead and adjust its behaviour. It's very helpful when we set a fast attack to make the compressor to react more promptly.
Multi-band compression is not very different from a standard compressor. In fact, all the parameters are the same, but it can split the signal in different ranges of frequencies so that each band will affect only that portion of the signal. This becomes very useful, for instance in mastering. A multi-band compressor, in fact, allow to correct certain dynamic problem on a specific range of frequencies leaving the rest of the mix unaltered.
Parallel compression (New York Compression):
Parallel compression is achieved by mixing the original and uncompressed signal with a heavily compressed version of the same signal. In order to create a copy of the original signal, we can just duplicate the channel or create a pre-fader send to a bus where we will apply the compression. Some plug-in also has a built-in "mix" knob which allows us to achieve parallel compression by blending the wet (processed) and the dry (unprocessed) signal.
Compression Practical Advises
In this session, I'd like to give you some practical advice on how to use a compressor.
Please take these as starting points and keep in mind that the setting of the parameters is strictly related to the material you are working on.
Generally speaking, a 3:1 ratio and 4dB of gain reduction is a good starting point for any audio source.
If you are looking for punch:
If you are looking for glue:
If you are looking for taming some peaks:
Ratio= 10:1, 20:1, inf.:1
Gain Reduction: probably 1-2db you want only catch the peaks
Classical parallel setting:
Some of the most famous compressors
Some of my favourite plug-ins:
(I try to keep this list updated)
IK Multimedia: 1176, LA2A, Dyna-Mu, Opto
Focusrite: Red 3
Slate Digital: FG Grey, FG Stress, FG 401, Audified U73B
Acustica Audio: Magenta
Waves: API2500, Renaissance Comp, Renaissance Vox, MV2