Beginner’s Guide To Audio Compression

The use of a compressor on an audio signal is meant to limit the dynamic range, the different between the loudest and softest parts. A good example of this would be a kick / snare drum, in the mix you might want the transient and low end of each drum hit to be consistent throughout the track. If the drummer lacks the natural consistency you’re looking for, you might have to turn to a compressor. Think of a compressor in terms of an automated fader for your audio signal, you could technically get the same effect by riding the volume for each track, but who has the time for that? This is where the audio compressor comes in!

A Brief History Of The Compressor

Like many other audio processing tools, we utilize in the world of recording and mixing, compressors began their life in radio. James F. Lawrence II, founder of the Teletronix Engineering Company in the 1960s, invited the LA-2A optical leveling amplifier. He served in the Navy during WWII as a radar operator and studied electrical engineering but his true passion lied elsewhere – radio and broadcasting. While working a position at KMGM in LA, he became frustrated with constantly having to manually ride the gain to keep a consistent level on air. Out of this frustration and from his engineering background and experience with military optical sensors, the idea of an optical leveling amplifier was born.

For all the compressors we have today, the idea remains mostly the same. The differences come from advancements in electronics and variances in additional functionality, attack / release times and coloration in terms of how the compressor is applied.

Compressor Functionality Breakdown

When we think of a modern day compressor, aside from a few exceptions, there will always be a standard set of controls: Threshold, Ratio, Attack, Release and Make-up Gain. These controls allow you to shape the audio signal as needed and outline the core functions of a compressor. Lets go over what each control does and how it is used.

Threshold

The threshold controls when the compressor begins to affect the incoming audio signal. If the audio is peaking at say -4db and the threshold is set to +5db, no compression will happen to the audio because it will never surpass the +5db level. If the threshold is set to -16db, compression will now begin affect the audio signal for every time it passes above -16db. Think of this control as a way to tell the compressor how hard is needs to work and how much gain reduction to apply. Speaking of gain reduction, this would be best described with the next control, ratio.

Ratio

The ratio works in tandem with the threshold control and tells the compressor how much the signal exceeding the threshold is attenuated or “turned down” by. Lets take the most common ratio setting of 4:1, this tells us that for every 4db that exceeds the threshold setting, the output signal will be attenuated to 1db above the threshold setting. For example, lets say you have an audio signal that is peaking at -4db and you set you threshold to -16db so that the compressor begins working on your audio signal. With a ratio setting of 4:1, the output audio level will be attenuated to a maximum peak output level of -13db, or 3db above the threshold. Lets take look at the math going on here, the incoming audio signal is peaking at -4db which ends up being 12db above the threshold setting. Our ratio setting of 4:1 tells us for every 4db above the threshold, attenuate the output to 1db above threshold. So, 4 goes into 12 3 times, this means the maximum peak output level is -16db + 3db which equals -13db. Lets say we change the ratio to 8:1 and leave all other setting the same. That would be 8 goes into 12 1.5 times, so -16db + 1.5db equals a maximum peak output level of -14.5db.

While it’s good to have a technical understanding of how compression ratio works, how do we use it? A good way to think about ratio is to break it up into 4 main types; light (1.5:1 to 3:1), medium (4:1 to 8:1), strong (10:1 – 18:1) and limiting (20:1 and above). We can use these guidelines to determine how hard to compress a given audio source. Keep in mind that the harder you compress an audio source, there will be less overall dynamics and a more pronounced sound of compression. This can lead to a flat, lifeless and/or pumping sound on some sources but could also help bring out quiet subtle details on others. The best thing to do is experiment with different settings on as many different audio sources as you can!

Now that we understand exactly how the ratio works with the threshold to attenuate the input audio level, we can move on to learning how to shape the ratio to affect the audio signal in different ways with attack and release!

Attack

Attack will control how long it takes the audio signal to reach full compression after passing the threshold setting. Attack times are normally measured in microseconds and milliseconds with settings between 20 microseconds and 5 milliseconds being considered faster and 10 to 150 milliseconds being slower. Attack will be focused on the transient, or initial large peak, of an audio source. A faster attack time will end up attenuating the transient while a slower attack time will allow more transient through. A good example of this would be to look at the audio wave form of a single snare hit, you will see a short pronounced transient peak at the very beginning which quickly drops off to a quieter and longer tail or “body” . The initial transient is the stick hitting the drum head, a very sharp loud sound which will peak higher on the meter and will stick out in a mix.

Release

Release works in the exact opposite way as attack, it controls how long it takes for the audio signal to go from its compressed state back to normal or uncompressed. Release times will be much longer than attack times and will be measured in milliseconds and seconds with 50 to 150 milliseconds being considered faster and 1, 2 and even 5 seconds being slow. Release will be focused on the tail or “body” of the audio signal which will include note sustain, tone, ring out, room reflections, ect..

Make-up Gain

Make-up gain is essentially an output volume knob that is used to match the volume of the compressed and uncompressed signal. A gain reduction meter in the form of a VU or LED lights will show you the exact amount of decibel level the input signal is being attenuated by. This meter, along with your ears, can be used to adjust the make-up gain for the output signal. Proper technique would be to always make sure the output signal matches the input signal, you can check this by frequently by-passing the compressor and listening for volume differences.

Using Attack & Release

These two controls are very useful because depending on source material, you may want to hear more or less of the transient compared to the tail or “body” of the source sound. Using the snare example again, it’s common to use a compressor to attenuate the transient and to bring the ring out and tone up in volume giving the snare hit more length, instead of it sounding like a quick smack or pop. This is accomplished by setting the compressor to fast attack and a fast release times which will allow the compression to come in right away affecting the transient then quickly going back to its uncompressed state leaving the body unaffected. Since the transient in this example would be peaking at a higher decibel level than the body, the result will be overall less volume. At this point we use make-up gain to bring the signal back to the level it was before, giving us an end product that has a more balanced transient to body volume level.