Compression:

The
Definitive
Guide for
Producers

Compression is one of the most

misunderstood and misused audio
effects in music production.

Some people argue it makes things

sound louder, some say it helps to
make things punchier.

All jokes aside, what is it actually

doing, and how can you harness its
power to make it work in your
tracks? There are so many
controls…

Don’t worry. In this definitive guide,

we’re going to lay everything out,
from the basics to the advanced.

Let’s look at how compression

works.
FR
EE

Want all the info from

this guide in a one-
page cheat sheet?
Download our FREE
Compression Cheat Sheet
below.

Download
Now

What is Compression
and Why Use It?
Before we dive into the complexities and
technicalities of how a compressor works, it’s
important that you understand it from an overall
perspective.

Otherwise, you’ll make the same mistake I did:

endlessly tweaking a compressor, hoping it
would fix my crappy songs.

Compression as a Tool
First of all, it’s important to understand that like
any audio effect, compression is simply a tool.

In fact, it’s an important tool that I would place in

the top 4 most crucial for mixing:

Volume
Panning/Stereo
EQ
Compression
It’s a tool that can be used well, or poorly.

Unfortunately, compression often ends up

becoming praised as the holy grail of a fat, full
and loud mix.

While it can contribute, many producers overlook

many other factors involved with mixing and
production as a whole, like:

Sound selection
Layering
Relative volume

How The Hell Does

Compression Work?
Compression is just volume changing over time,
really fast.

Read that again.

It’s just volume automation.

How? It reacts to the incoming audio and turns it

down when it gets too loud, and turns back up
when it’s quiet again.

Pretty simple right?

This reduces what is known as the dynamic

range of a sound.

The dynamic range is simply the difference

between the quiet and loud parts of a sound.
Compression reduces this, making it more even
over time.

But why would you want to do something like

that?

Why Should I Use

Compression?
Some audio material tends to be very dynamic,
meaning there is a large gap between the quiet
parts and the loud parts of the sound.

This can be good, but it means that certain

elements in a track might not poke out the whole
time, burying them in the mix.

Maybe a snare punches through too loud and

the rest gets buried right after. Perhaps certain
guitar strums don’t quite cut through. Maybe
your voice sounds muffled in parts, but clear in
others.

The process of evening out instruments and

sounds is very important for recorded music, as
listeners want a consistent listening experience.

This is especially important in 2020, where

music can be listened to in all sorts of
environments – at home chilling, when walking to
work, in the car, with your friends at the park, in
the club etc.

Some of these environments may have different

sounds going on, so the music needs to poke
through in order to keep up.

Most of the time, compression can help us

achieve this effect.

So let’s take a look at the basics so we can start

to use this great tool.

Basics of a Compressor
To understand how a compressor works, you
have to know each of the main controls:

threshold
ratio
attack
release
gain reduction
makeup gain
Some compressors will deviate from these, some
may actually exclude these completely. But we’ll
start with these basic controls and then dive into
less common features.

Threshold
The threshold is the ‘when’ of the compressor.

It says ‘it’s too loud now, time to turn it down!’

This control is measured in dBFS, with 0dB being

the loudest and -∞dB being silence.

Where you should set the threshold is very

dependant on the incoming signal, and it can
help to have some sort of visual on the signal to
see where it’s peaking at.

Ratio
The ratio is the ‘how much’ of a compressor.

It determines when compression is engaged,

how much the signal is turned down by.

This is measured as a ratio of dB from input to

output.

For example, a ratio of 2:1 will turn a signal that

is 1dB over the threshold down by 0.5dB.

Because for every 2dB that crosses the

threshold on the input, it will output only 1dB
over the threshold.

But a ratio of 4:1 will turn that signal down by

0.75dB, leaving only 0.25dB over the threshold
left.

Because for every 4dB that crosses the

threshold on the input, it will output only 1dB
over the threshold.

This can be a confusing concept, but higher

ratios = more compression.

Ratios of around 2:1 and 3:1 are great for general

dynamic control, and upwards of 4:1 are closer
to limiting and squashing type effects.

Combining Ratio and Threshold

Ratio and Threshold are the two key players in
compression, and how they are set in relation to
eachother can make or break the type of sound
you are trying to achieve.

Most people take one of two approaches when

setting ratio and threshold. High ratio and low
threshold, or low ratio and high threshold. The
first generally applies less-frequent
compression, but it’s heavy once it engages.
High
High

The latter applies a subtle amount of

compression more consistently.

Either method can be fine, and you don’t have to

always adhere to either rule (low ratio and
threshold can be fine for very subtle control). It
ultimately depends on the type of sound you are
going for, although many would argue that brutal
compression sounds unnatural, even when
infrequent.

Attack
Both attack and release determine the speed at
which the compressor works. At 0ms, both
controls work instantly, but this doesn’t sound
very musical and often introduces distortion.

Specifically, the attack determines how long the

compressor takes to engage.
Short
Long

Longer attacks take more time to kick in, usually

resulting in less compresssion.
A fast attack makes a sound squashed and
controlled (great for audio spikes or overly-
dynamic sounds), whereas a slow attack gives
the start of a sound time to breathe (great for
percussion).

Listen to the snare below. The first has a short

attack, the second has a longer attack, as
pictured above.

0:00 -0:00

0:00 -0:00

Note that both attack and release are gradual

controls. An attack time of 50ms means that it
will fade in the compression over that 50ms, not
immediately after 50ms has passed.

Release
In conjunction with the Attack, the Release
determines how long the compressor takes to
disengage, once the audio has passed back
below the threshold.
Short
Long

Like with attack, a fast release can sound

squashed and gives a pumping effect to
compression (which can be desired).

However, a slow release can sound smoother

and slower, which is great for more subtle
applications.

Listen to the two examples of short and long

release on this vocal sample. The first sounds
very pumping, whereas the second sounds
smoother but more compressed.

0:00 -0:02

0:00 -0:02

Like with the attack, the release will fade back to

normal over the set time. It doesn’t wait for say
200ms and then suddenly stop compressing.
This is how a compressor maintains a
transparent and natural sound.

Gain Reduction
Not so much as a control, but gain reduction is
an important measurement to keep an eye on.

Why? It measures how much compression is

taking place at any given moment.

Gain Reduction Graphic

Crank up the ratio or threshold and you’ll see the

gain reduction meter increase as a result, so it’s
great to use in conjunction with those controls.

Makeup Gain
The last of the basic controls is makeup gain.

This determines how much the signal will be

brought up in volume after compression has
been applied.

Makeup Gain Graphic

A good practice here is to match the new peak

level to the old peak level. Even though they
have the same theoretical ‘max’ loudness peak,
one sounds fuller and louder.

Recommended: Mixing For Producers

Getting More Advanced

So we’ve covered the main controls of a
compressor, and you’ll find those on 99% of
plugins. But each compressor starts to include a
variety of the features below, so let’s see what
else can crop up.

Knee
Moving onto more advanced controls, we have
the knee.

Knee, measured in dB, can be powerful if used

correctly, even if it’s a subtle effect.

It works by ‘blurring’ the threshold. Instead of

engaging at a very specific point, the ratio
creeps up from 1:1 (no compression) to the
actually defined ratio over a set amount of dB.

Knee Graphic

Say your threshold is -18dB and your ratio is 4:1.

As the audio approaches -18dB, with a knee of
3dB the compressor will actually start engaging
1.5dB earlier (knee blurs both ways, 1.5dB before
the threshold and 1.5dB after).

So as your audio hits -19.5dB, you’ll start to hear

more subtle compression, and as it moves to
-16.5dB, the full ratio will be engaged.

Input Gain
In addition to the makeup gain control, some
compressor will include an input gain knob.

Input Gain Graphic

This allows you to drive the signal into the

compressor, creating a greater sense of
compression.

It also helps if the signal you’re working with is

too loud or quiet, as you can set it to an ideal
volume to work with.

Lookahead
Compressors, although they are useful tools, are
unfortunately quite dumb, and are reactive as
the audio comes in.

Sometimes, this means the compression doesn’t

engage fast enough and clicks and pops still get
through when large dynamic changes occur.

Lookahead Graphic

Lookahead solves this issue by adding latency in

order to see the signal for a bit of time before
outputting the compression. Even something
small like 1ms can make a large difference.

Range
Say you’ve set your ratio and threshold to a
position where you are consistently getting
between 3dB and 6dB of gain reduction. You like
the effect it has sonically, but it’s a bit too
intense.

Range scales back the compression by reducing

the amount of dynamic range the compressor is
‘allowed’ to look at. Think of it as ‘relaxing’ the
ratio control.

Glue Compressor - Range

Ableton Live’s Glue compressor limiting the

amount of gain reduction to 5dB

By default, the compressor could reduce the

signal down to -∞dB if possible, but bringing the
range up can control more intense gain
reduction.

Dry/Wet
This is a more modern feature on digital/software
compressors that allows you to easily blend both
the wet and dry signal to dial back the amount of
compression.

Dry/Wet Graphic

This is also known as parallel compression, as

you are mixing the uncompressed signal with the
compressed signal.

Before software compressors, a compressor was

used on a send/return and was blended in with
the original signal, but now you can do it all on
the same effect.

Peak & RMS

Another more advanced control is peak/RMS
switching.

Some software compressors will offer this

feature, while analogue or modelled
compressors will rely on the circuit type to
perform either type of compression (some
analogue compressors would be closer to RMS,
some to peak etc.)

Ableton Live Compressor Peak or RMS

Peak compression acts on the momentary audio

as it is, whereas RMS compression acts on the
average loudness of the audio (closer to how
humans hear sound).

Peak mode is great for precise compression

(getting a snare drum to punch) whereas RMS
can be more ideal for smoother, more musical
compression (compressing a whole mix).

Curve Type
Certain compressors will allow you to choose
between different shapes of curves. It’s not a
very common control, but can be useful to know.

The most common two are Linear and

Logarithmic.

Logarithmic tends to be how most audio is

compressed, as it sounds natural to us.

Linear compresses in a straight line and sounds

precise but very digital, especially when heavy
compression is being applied.

Some other compressors (FL Studio’s Fruity

Limiter comes to mind) have multiple different
release curve types, as well as attack curves.

FL Studio Fruity Limiter Release Curves