Introduction to Music Production 4 – Compression
Compression, in audio, is used to shorten (compress) the dynamic range of a sound. The dynamic range of a sound is measured in dB.
Threshold: The value in which will trigger the compressor to engage on in the event the incoming audio signal surpasses the threshold. Measured in dB.
Attack: The amount of time an audio signal must stay above the threshold to engage the compressor. Measured in Milliseconds.
Release: The amount of time an audio signal must stay below the threshold to disengage the compressor. Measured in milliseconds.
Ratio: The degree to which an audio signal is compressed. Higher ratio = harder compression. A ratio of 2:1 means that any audio that is being compressed, for every 2 dB above the threshold, the range will be reduced to 1dB. A 4:1 ratio means for every 4 dB a sound is above the threshold, its range will be reduced to 1 dB (so +8 dB above the threshold would be 2 dB after compression)
Knee: A value determining the smoothness around the point at the threshold. Lower knee = harder angle, higher knee = softer curve. Measured in dB.
Output Gain (Makeup Gain): This is used as a volume fader at the end of the compressor in the event the output is too loud or soft. Measured in dB.
Sidechain/Key input: This is a function that lets another instrument trigger the threshold of the compressor. More on this later!
Compression can achieve multiple outcomes depending on the settings. With a very quick attack, a lower threshold, and a high ratio, you can ‘squash’ an input signal so that the dynamic range is much smaller. With an attack of about 2-3ms, medium threshold, and short release, you can accentuate the attack of percussion sounds. With a quick attack, a medium threshold, and a short release, you can un-accentuate the attack of a percussion sound. It can smooth out an overly-dynamic vocalist for a better studio production.
Here is an example of a kick drum with compression settings to help accentuate the attack. In this case, the attack itself is not being processed by the compressor, but the decay is, creating a larger dynamic distance between the attack and the decay. This means the decay is being compressed to a lower dynamic range (in this case, creating a lower volume), while the attack stays the same.
You will first hear the unprocessed (dry) kick, then the processed kick.
Here is an example of a compressor with a quick attack and a longer release to keep a crash cymbal sustained at a higher volume for longer.
First you will hear the unprocessed (dry) crash, then the processed crash.
When compressing vocals and instruments, especially live recordings, be aware of background noise. If there is a lot of background noise, you may not want to compress the signal, and instead use volume automation to achieve more steady dynamics (volume changes).
Compression is a topic that is better practiced than read about. Go place compressors on instruments and turn knobs!
Sidechain compression is a concept in which you use an external trigger to engage the compressor on the audio that is being processed. This is commonly used to make room for the bass drum, as well as other instruments, when frequency masking can occur. More dramatic examples of sidechain compression are commonly found in all electronic music and hip hop. The concept of sidechain compression is also called ‘ducking,’ because it ‘ducks’ away an instrument from the trigger sound.
Here is an example of a kick and a bass, where the bass is sidechained to the kick. You will first hear the kick and bass with no processing, then you will hear the ducking that occurs from sidechain compression.
Compression is one of the more misunderstood audio processing methods, but with practice and listening, it is a crucial tool in production, though it is not a necessary effect on every single sound.