Last week, we explored basic EQ principals and nudge your recording project into the mixing phase. Now, as we near the close of this series, it’s time to add some compression.
Compression is one of the most misunderstood elements of audio, so let’s clear up a common misunderstanding right out of the gate. There are two different types of audio compression: dynamic range compression and data rate compression.
This article is going to cover dynamic range compression, which affects the amplitude of a waveform. Data rate compression, on the other hand, affects the size of an audio file.
What Is Compression?
Think of it like this... You know that scene in Jurassic Park where the kids are hiding in the car, and they see the water rippling in the glass as the T-Rex approaches? That’s one of the quietest scenes in the movie. The T-Rex shows up and lets out an ear-piercing roar. That’s one of the loudest scenes in the movie, and it feels really loud because the scene right before it was so quiet. That’s the power of a large dynamic range.
If you happen to be watching on regular cable, you’re greeted with those screaming advertisements as soon as a commercial starts. It feels like the commercials are even louder than the T-Rex, and that’s because it never got quiet. That’s the power of a small dynamic range.
Compression controls the dynamic range of an audio signal.
Dynamics are the fluctuations from quiet to loud within a performance, and dynamic range is the difference between the loudest part of a song and the quietest.
Compression squashes the loudest parts of a signal, allowing us to make the it denser. It makes loud sounds quieter, and quiet sounds louder.
What to Compress
A common mistake new engineers make is slapping a compressor on every channel simply because they think they’re “supposed to." Adding plugins to channels can feel like progress, but if you’re not using the tools correctly, you could actually be doing more harm than good.
It’s really important to know why you want to compress a signal—it tells you which compressor to use and how to set it. Here are some common reasons engineers use compression.
Controlling Dynamics: The primary job of a compressor is to control dynamics. You can do that by compressing the transients (or peaks) of a signal to make levels more consistent. This can also be used to prevent clipping when tracking or even during mixing and mastering.
Note: Transients are the short spikes in volume from the initial attack of a sound. Snare hits, vocal consonants, and the moment the pick hits the guitar string are all great examples of transients.
Shaping Dynamics: You’ve probably heard the word “punch" thrown around a lot when talking about compression. Compression allows you to change the envelope of sound, which lets you control the initial punch of a sound.
Tone: Every single compressor has it’s own unique tone that it imparts on the signal. Some compressors are great for fattening up a bass guitar, while others are better at making a snare drum jump out of the speakers. Some compressors are specialized to focus on a particular frequency range, like a de-esser or a multi-band compressor.
Proximity: Compressing a signal can “bring it forward" in the mix and make it sound like that instrument is closer to you than the others. Conversely, it can also “push it back" further in the mix.
Since compression makes loud sounds quieter and quiet sounds louder, it has the added bonus of increasing perceived intensity. For instance, if you compress a room mic from a drum recording session, you can make it sound as if the drummer was playing louder than they actually were by making the quiet reverberations feel almost as loud as the drum hits themselves.
Compressor Controls
Now that you understand why to use a compressor, let’s talk about how to use a compressor.
Threshold tells the compressor when to start compressing. If a signal gets too loud and goes over the threshold, the compressor kicks in and lowers the volume. This setting will vary greatly from track to track (even more than the others).
Ratio tells the compressor how hard to compress the signal. Low ratios (like 2:1) will be more subtle, while higher ratios (10:1 and above) are much more noticeable.
Too much can suck the life out of a performance and push it back in the mix."
Attack tells the compressor how quickly to compress the signal. A slow attack allows more the impact of the transient to come through, which helps add punch but doesn’t do much for taming peaks and controlling dynamics. A fast attack actually shaves off the initial transient, which can help tighten up a performance. However, too much can suck the life out of a performance and push it back in the mix.
Release tells the compressor how long to compress the signal. A slow release sounds smoother and allows more dynamic control but can also suck the life out of a performance. A fast release has a more aggressive and gritty sound that brings up the initial sustain and increases perceived lounges, but too much can cause pumping (or audible compression).
Generally speaking, faster attack and release times will give you more aggression, grit, and loudness, while slower times sound smoother.
Make-up gain tells the compressor how much gain to add back to the signal so it’s as loud as it was before you compressed it. A compressors job is to reduce gain, and our ears are hard-wired to believe that louder is better, so the only way to accurately compare before and after compression is to use make-up gain to add back however much volume you've knocked off with the compressor.
Types of Compressors
Some compressors are better at certain jobs than others. There are four main types of compressors, each with their own pros and cons:
Optical is one of the oldest and crudest forms of compression. It uses light to trigger gain reduction, which gives it a very slow attack time. It’s surprisingly musical, and great for smoothing out inconsistent levels and fattening up sounds. The LA-2A is one of the most popular optical compressors.
Tube / Variable-Mu uses a vacuum tube to apply compression. It doesn’t have ratio control, and instead, ratio is increased as input level is increased. It has unique tone and is great for glueing instruments together. Variable-Mu is a proprietary technology developed by Manley for their Fairchild with “program dependant" settings, meaning the ratio and attack/release times change depending on the incoming signal. The Fairchild 660 and the Manley Vari-Mu are some of the most popular vintage tube compressors.
FET uses a “Field Effect Transistor" to control gain. It was developed as a faster, snappier version of tube designs. Very colorful and great for increasing intensity and bringing instruments forward in a mix. The most popular vintage FET compressor is the Urei 1176.
VCA is a modern technical design that uses a voltage-controlled amplifier to control gain. It can be smooth or fast and punchy. Attack and release times range from very slow to very fast, and it’s great for adding punch and excitement. One of the most popular vintage VCA compressors is the dbx 160. One of the most common VCA compressors comes from the SSL channel strip.
The Empirical Labs Distressor has also been gaining a lot of popularity in recent years for its ability to emulate the tone of optical, tube, FET, and VCA compressor types.
Setting a Compressor
- First, turn the threshold knob all the way down to 0dB and select an appropriate ratio.
- Then, set your attack time as slow as it will go and your release time as fast as it will go.
- Slowly turn down the threshold until you start to see the amount of compression you want (~3-6dB is a common starting point).
- Tweak the attack time until the right amount of transient is getting through.
- Tweak the release time so the compressor “breathes" with the track, which means that the compressor releases just before the next transient hits.
The attack and release times will vary from greatly from track to track, but when done correctly, it gives the song a natural, organic sound like all of the instruments are working together.
Other Dynamics Processors
Sometimes you may find that a compressor just can’t control a sound the way you want. There are several other types of dynamics processors, each with their own job.
Limiters: Technically, any compressor with a ratio over 10:1 is considered a limiter, but most engineers are using it loosely to refer to a “brick-wall-limiter." These limiters have a ratio of infinity:1, meaning it’s literally impossible to clip if set correctly. Typically, it’s used to increase overall volume when mastering.
De-Essers: They do exactly what they sound like: remove “ess" sounds. Technically, they’re just a single-band compressors that focus exclusively on high frequencies to remove sibilance from vocals and overly bright instrument recordings.
Sometimes, a single compressor isn’t enough...That’s where multi-band compressors come in."
Multi-Band Compressors: Sometimes, a single compressor isn’t enough. Sometimes, you need to compress each frequency range differently (typically when mastering). That’s where multi-band compressors come in. Typically featuring 3-7 bands with their own threshold, ratio, and attack/release times to control the entire frequency spectrum. Also used in mid/side processors.
Gates: You know those electronic sliding doors on Star Trek at the supermarket? That’s basically what a gate does for your audio. They have the same settings as compressors, but instead of reducing the gain of a signal when it goes over the threshold, they mute the signal when it falls below it. Basically, it’s the "you must be this tall to ride" sign of the audio world. If you’re not loud enough, you get muted entirely. Typically used to remove bleed in close mic’d drums.
Expanders: Honestly, these are just fancy gates. Instead of muting the signal when it falls below the threshold, it reduces the gain by a certain amount. It’s a slightly more subtle form of gating, and it’s typically engaged by pushing a button on the gate module itself.
It’s All In Your Head
Now that you know the basics of dynamics processing, it’s time to get practicing Just remember that plugins don’t always equal progress. You don’t need to put a compressor or a dynamics processor on every channel. Think about why you want to compress an instrument and how you can achieve the desired result before slapping on your favorite vintage compressor emulation and loading a preset.
It’s easier to make a track sound worse with compression than it is to make it sound better. As always, just remember to use your ears, and trust your gut.
Editor's Note: Check out every installment in Brad Pack's series on home recording.
- Home Recording Basics I: Setting Up Your Studio with Monitors and Acoustic Treatment
- Home Recording Basics II: Choosing Your First Audio Interface and DAW
- Home Recording Basics III: A Guide to Microphone Types and Placement
- Home Recording Basics IV: How to Record Acoustic Guitar and Vocals
- Home Recording Basics V: Mic'ing a Drum Kit From Start to Finish
- Home Recording Basics VI: How to Mic Your Electric Guitar
- Home Recording Basics Part VII: Recording Vocals in Your Bedroom
- Home Recording Basics Part VIII: Basics of MIDI and MIDI Controllers
- Home Recording Basics Part IX: How to Prep Your Tracks for Mixing
- Home Recording Basics Part X: Equalization
- Home Recording Basics Part XI: Compression and Dynamics Processing
Learn more about how to record music at home on our Basics of Home Recording homepage.