Compression is one of the most commonly used effects in music production. It’s also one of the hardest to understand properly. Other effects, like reverb, delay and chorus make an obvious difference to the sound, but with compression it can be harder to hear what’s going on.
The way we are using the term “compression” here relates to dynamic range reduction. This means that the difference between the loud parts and the quiet parts of a track or audio signal is reduced. A compressor, either hardware or software, is used to to do produce this compression.
This doesn’t sound very exciting, but compression is important for enhancing the sound of the production when mixing and mastering, and even to produce special effects that have become essential in some styles of music.
Here we have a look at what compressors are, what they can do and a little bit about how compressors and compression has developed over the years.
What Is a Compressor?
The explanation of a compressor that helped me the most is that it is an automatic volume control. When the sound gets too loud the volume is turned down a little, and when the sound gets quieter the volume is turned back up. This volume adjustment happens very quickly, within a few milliseconds.
The automatic adjustment means that you can have the the volume high enough to hear the quieter parts of a track, but have it turned down so that it doesn’t cause distortion when the loudest parts are playing.
So a compressor lets you control the peaks in an audio recording. It lets you control the volume of louder signals without affecting the quieter signals. With the volume of the louder sections reduced you can turn the whole thing up. This has the overall effect of making the track sound louder overall, and usually makes it sound better too.
Technically, it’s the gain that’s being adjusted by the compressor. Gain is the increase in power of the audio signal, usually measured in decibels (dB). Since it’s just the basics of compression that we’re looking at here we won’t worry too much about using the exact terminology. Purists might not like that, but we’re just trying to provide some helpful general information here.
The First Compressors
Since a simple compressor is really just an automatic volume control, it’s not surprising that early compressors were an attempt to catch peaks in a signal and avoid it going over a particular level. The earliest attempts at this were aimed at radio broadcasting to prevent distortion, or even failure, of the signal being broadcast due to overloading.
The earliest compressor was called a “leveling amplifier”, and it was designed in the 1950s to reduce the level, or gain, of a signal when it went over a specific threshold. Electronics has come a long way since the 1950s, and the way in which the early leveling amplifier compressor worked seems very old fashioned now.
As the level of the audio signal entering the device increased it made a bulb glow brighter. The light from the bulb shone on a light-dependent resistor. As the light got brighter the resistance in the resistor increased, which reduced the strength of the signal passing through it. This reduced the level, or gain, of the signal.
When the level of the signal entering the device was lower the bulb glowed less brightly, which reduced the resistance of the light-dependent resistor and the level, or gain, of the signal passing through was increased.
If you are familiar compressors you may have heard of the Teletronix (Universal Audio) LA-2A, which is a hardware compressor that usually comes top in polls about which is the best compressor you can get.
I had a real ah-ha moment reading about this when I discovered that the “LA” in the name of the compressor stands for “leveling amplifier”. I had assumed that LA stood for Los Angeles, as in “if you use this compressor you will get the LA studio sound”. Just goes to show what assumptions are good for.
This unit, which costs the best part of $4,000 dollars to buy, is the descendent of these original compressors and it still uses the electro-optical circuitry described above.
I don’t have enough money to be able to afford a hardware LA-2A, but I do have a software plugin version, the Waves Audio CLA-2A, which you can see below.
Controls on a Typical Compressor
The Teletronix LA-2A is known for its two-knob compression system, but modern compressors used in music production usually have six basic controls that let you do most things that you need to do.
Many compressors have more controls, and a few have less, but we’ll look at the six main controls first to see how they work and cover any additional ones later.
The five main controls are: threshold, ratio, attack, release, knee, make up gain.
The threshold control lets you adjust the level at which the compression of the signal will begin. When the level of the signal entering the compressor goes above this threshold it will be compressed.
The threshold level is usually measured in decibels, usually shortened to dB. The concept of the decibel is one of the most confusing things when working with audio.
The scale for the threshold control typically goes from around -60dB up to 0dB. If the threshold is set at -6dB, for example, then the signal would start being compressed every time it reached that level.
Having the threshold higher than the loudest part of a track means the compressor will never be triggered. Lowering the threshold means that more of the signal will be compressed.
The ratio control lets you adjust how much compression will be applied to the signal. The full name for this control is Compression Ratio, but that’s usually too much to fit next to the knob or slider on the device.
The ratio refers to the how much the level of signal coming into the device is reduced before it is output from the device.
As the name suggests, the ratio levels are expressed as ratios, and the values typically go from 1:1 (no compression) to 30:1 (a lot of compression). The first number refers to the level of the incoming signal and the second number relates to the level of the signal coming out of the compressor.
Remember, it’s just the part of the signal that exceeds the threshold that has been set that is compressed. The audio below this level is unaffected.
Here’s an example. If the ratio of the compressor is set at 4:1, then for every 4dB the level of the incoming signal exceeds the threshold level the output signal is reduced to 1dB above the threshold. With the same ratio setting (4:1) a signal that is 8dB above the threshold will be reduced to 2dB above the threshold.
If the ratio is set at a high level then the compressor device becomes a limiter. A limiter cuts the signal off at the threshold so that the signal output doesn’t exceed the threshold level at all.
Strictly speaking a limiter should have a ratio of infinity:1, so that the signal is squashed down to the threshold level, but in practice a compressor with a ratio of 10:1 or more is considered to be acting as a limiter.
The attack time control lets you adjust how quickly the compressor will start reducing the level of the signal once it exceeds the threshold. This can be used to fine-tune the compression that is produced.
For example, with drum sounds you might want the initial attack of the drum hit to be left alone (even though it exceeds the threshold) but to compress the rest of the sound of the drum. This can be done by increasing the attack time so that the compressor doesn’t start working until after the initial drum hit sound has passed.
The release time control lets you adjust how long it takes for the device to stop compressing the signal after it has gone back down below the threshold level. The delay before releasing compression is needed to help provide a natural sound, which would not be possible if the compressor just stopped as soon as the level wen’t below the threshold.
In some types of music, a short release time is used to produce specific effects. This type of thing is particularly electronic dance music, and an example is the use of a short release time to produce the popular pumping “pumping” effect.
As mentioned above, sometimes you want the compressor to produce a natural sound, while other times you want to produce something a little more dramatic and noticeable.
The knee control lets you adjust how dramatically the compressor starts to compress the signal once it has been activated by the signal exceeding the threshold. The two main options are “hard-knee” and “soft-knee”.
With hard-knee compression, the compression immediately starts being applied as soon as the compressor is activated.
With soft-knee compression the compressor is able to monitor the incoming signal and compression is gradually applied as the signal approaches, and then exceeds, the threshold.
The compressor starts compressing at a ration less than what has been set, and only reaches the set compression ratio once the signal has exceeded the threshold.
Soft-knee compression lets you achieve a less noticeable, more transparent, compression. Hard-knee compression can be used to make the compression more noticeable, so that particular audio effects can be achieved.
The last of the six common compressor controls is Make Up Gain. Since the peaks of the audio signal have been compressed, the overall volume of the audio will have been reduced.
Make up gain lets you turn the overall level of the sound back up to compensate for the compression. Even though the level of the audio signal has only been put back to where it was before compression, you will be turning up the quieter parts too so the overall effect is an increase in loudness.
Multiband vs Single Band Compressors
Most compressors compress the sound more or less equally right across the range of audio frequencies, i.e. from the lowest to the highest pitched sounds.
Some compressors are able to divide the audio signal up into a number of separate frequency bands. This means that some frequencies can be compressed more (or less) than others, which lets you compress each band as required rather than having to compress the whole signal as one. These are called Multiband Compressors. The image below shows the separate frequency bands on the Cubase Multiband Compressor.
You have probably heard of “De-essers”, or heard about vocal recordings being “de-essed”. This is a process that lets you reduce the amount of sibilance in a recording. Sibilance is a harsh high frequency sound that is usually produced by the letter “s” being sung or spoken (although other letters produce a similar effect).
A de-esser is a type of compressor that is set to compress the sound at the particular frequency where the sibilance occurs. This reduces its level in the recording, which makes the overall sound more pleasant.
Types of Audio Compressor
The original hardware compressors used a variety of different technologies to produced compression of audio signals. Compression of the signal results in dynamic range reduction.
Some software compressors are designed to emulate specific hardware compressors. Some of these software compressors enable you to choose between the different styles of compression all within one device.
Examples of the original hardware compressor technology, now emulated in software devices, include valve (or tube) compressors, optical compressors, FET compressors and VCA compressors.
Below is an image of the compressor that comes with Apple’s Logic Pro X. It has buttons along the top that let you switch between different styles of compression.
The first compressors that we looked at above were optical compressors. These hardware devices contain a bulb that gets brighter as the audio signal gets higher.
The light shines onto a light-dependent resister, which increases its resistance as the light gets brighter. This increase in resistance reduces the level of the audio signal.
Since there is a short delay, usually called latency, before the signal is compressed, optical compressors naturally produced soft-knee compression
As mentioned earlier, the most well known example of this type of optical compressor is the Teletronix LA-2A unit. Several software emulations of this classic compressor are available from companies like Waves, Native Instruments and Universal Audio, who also produce the hardware LA-2A device.
Tube compressors use valves (tubes) to produce compression. Like other electronic devices that use valves, e.g., guitar amplifers, tube compressors produce a warm, sometimes described as “lush”, sound. Like valve guitar amplifiers, tube compressors add something to the sound in addition to providing compression of the audio signal.
Tube compressors often have fewer controls than other types of compressors since things like ratio are handled automatically, with the valves compressing the signal more as the level gets higher.
Well-known example of tube compressors are the Manley Variable Mu and the Fairchild 670 compressors. Some software compressors aim to recreate the warmth produced by the valve technology used in hardware tube compressors.
FET is short for “field-effect transistor”. As you probably know, transistors can produce similar effects to valves in a much smaller, and less expensive, package. Devices using transistors rather than valves were often referred to as “solid-state” devices.
One of the advantages of using transistors to produce compression of the signal was that they are able to begin compression of the signal very quickly. This meant that compressors could have a much shorter attack time than was possible with tube (using valves) or optical compressors.
VCA is short for “voltage controlled amplifier”. Compressors using the VCA design are probably the most widely-used devices for general compression purposes. The are able to provide less noticeable (more transparent) compression of an audio signal.
VCA compressors work by using the level of the input signal to control the amount of compression applied. So the higher the level of the input signal, the more compression is applied to the audio signal before it is output from the device.
How Is a Compressor Used in Music Production?
Of all the effects that can be applied to audio, it’s probably harder to see how compression can be used effectively. People often talk about compressors being used to make a track louder, but it can be a lot more subtle than that.
When looking at compression a distinction is usually made between compression of individual audio tracks in a piece of music and compression of the entire piece (which is often referred to as “the track”, which can get confusing).
Compression on Individual Tracks
Probably the simplest way to use compression on individual tracks in a piece of music is to give them “more punch”. This is most commonly done with drums, particularly the kick drum, but compression can be used on any track in a piece to control peaks in the audio.
This helps to increase the overall level of an individual track and generally “glue” all the tracks together since none of them are dramatically jumping up and down in volume. One of the compressors that comes with Ableton Live is called the “Glue Compressor” because of its ability to do this.
Compression on the Overall Piece of Music
When compression is applied to the overall piece of music the compressor device is usually inserted into the stereo output channel of the recording.
The aim is usually to control the peaks of the louder sections of the final audio signal so that the overall level of the signal can be increased, or made louder.
One of the reasons this increase in loudness is needed right at the end of the production process is that it helps the audio to sound good on almost any type of speaker that the listener might use.
One very popular use of a compressor is for sidechain compression. This is where the compression is triggered by an audio signal separate from the one that is being compressed.
This was originally used by radio DJs so that when the started speaking into the microphone compression was triggered on the music playing so that it was turned down and the DJ could be heard over it.
This is commonly used where instruments like drums and the bass might clash with each other within particular frequency ranges. If the sound of the kick drum is used to trigger compression of the bass then every time the kick drum sounds the volume of the bass will be turned down a little. This produces “space” for the kick drum, which can improve the overall sound.
A more extreme version of sidechain compression is used in dance music to produce a pumping effect. Again the kick drum is used to trigger the compressor on other tracks, but in this case more extreme compression with a very short release time is used to produce a “sucking” or pumping effect.
Compressors in Audio Production
Compression can be used to subtly enhance the sound of your audio productions, or it can be used as a creative tool. It takes quite a lot of trial and error to understand how it works and what a compressor can do, but it’s well worth the effort.