The range of effects you could use in a studio are pretty much unlimited, but there are a couple that are genuinely essential. Compression is one of these effects, whether it’s software-based compression or a hefty chunk of ‘real’ hardware. But what exactly is compression? And how do you actually use it? This blog will try to answer these questions and more.
- Compression in a Nutshell
- The Finger is on the Button
- Basic Functions
- Attack and Release
- More Advanced: The Sidechain Explained
- Compression: A More Detailed Picture
- Compressor Terminology
- What’s a Limiter, Then?
- Practical Problems
- More Extensive Compressors
- More Creative Compression
- Compression Flavours
- Character Building
- FET Audio!
- Delta-Mu Compressors
- See Also…
Compression in a Nutshell
If this is your first encounter with the world of compression, then reading this little introductory snippet might be enough for you. If you’re feeling a little brave or you’re already familiar with the basics, then feel free to skip on over the more advanced section, ‘Compression: A More Detailed Picture’. Either way, enjoy!
The Finger is on the Button
Believe it or not but before the emergence of compression, the same service was provided by a humble technician with lightning-fast reflexes sitting at the volume knob of every radio station in the world. To make sure that radio broadcasts actually sounded good, the dynamic range of the music being played had to be kept to a minimum. Otherwise, quieter sections would get lost in noise and static while louder parts would be rendered unrecognisable by distortion. In the ‘50s, the quick-witted technician was replaced by the very first dynamic range compressors and it wasn’t too long before music producers realised that this hardware could do much more than automatically turn the volume up and down. They noticed an interesting side effect – an effect that somehow gave the effected music more punch … more focus … and pumped up the volume!
While compressors come in all shapes and sizes, the concept usually remains the same. The signal comes in via a line-input, a clever detector tracks the amplitude of the signal, and depending on the settings, the compressor adjusts the volume here and there and sends the resulting signal through the line-output. Compressors are often found sitting between a microphone preamp and some kind of recorder, or in the effects-loop of a channel on a mixer. The difference between most compressors lies in how much the parameter can be adjusted. Almost every compressor you might lay your hands on will have adjustable threshold. Here, you can set the audio level at which the compression actually kicks in. And, because the signal is attenuated (or slightly weakened) by its journey through the compressor, you’ll usually get a ‘make-up gain’ function to compensate for the loss and give it a little boost back up to the original level.
Attack and Release
With more extensive compressors, you can also asjust the speed at which the compression kicks in once the set threshold has been reached, and how long it continues to effect the signal thereafter. These functions are referred to (respectively) as ‘attack’ and ‘release’ and have a massive influence on the punch and rhythmic character of the source audio signal. With a fast attack, peaks in the signal are absorbed immediately, making it sound more soft and round. With a slow attack, the signal is only pulled back after any initial peaks – and the reduction is very brief. This actually emphasises any peaks – making it the perfect setting when recording something like drums. Using a short release time, the audio level is very quickly restored after it’s gone over the threshold, so any natural reverb in the recording booth, or the sustain from the instrument itself is maintained and in some cases, even amplified. So, if there’s any unwanted reverb caused by the space you’re recording in, you can opt for a faster release time.
The control panels of more extensive compressors usually also offer control over the amount of applied attenuation once the signal peaks over the threshold. This is called the ratio since it’s literally expressed as such: e.g 2:1. The higher the first number, the harder the compressor works. In fact, a Limiter is actually nothing more than an extremely hard-working compressor, or an extremely high ratio compressor (with a ratio of ‘infinite:1’). This means that no compression threshold is even set. Instead, any peaks are blocked completely.
More Advanced: The Sidechain Explained
How much work a compressor actually does relies on the signal it ‘hears’. Usually, the signal that the compressor ‘hears’ is the same signal that it’s effecting. But this is not always the case. While it may sound a bit weird, you can make a compressor ‘listen’ to a different signal than the one sent through the line-input. This sneaky technique is referred to as side-chaining, and is used a lot when producing electronic dance music. This technique allows the producer to compress a syntheiszer part while the compressor actually processes the kick. This has the effect of making the synth part sound like it’s being ‘pushed’ away at every strike of the kick, resulting in that essential ‘pumping’ feel.
Compression: A More Detailed Picture
By using compression, you’re editing the dynamic range of a signal. This way, very specific quiet or loud parts of a signal can be turned up or turned down. This is how it works: the compressor checks the audio signal and initially allows it to pass through its channel unhindered. However, if the compressor detects peaks of a certain loudness in the signal, then it leaps into action and narrows its channel to literally squeeze and ‘compress’ the signal until it’s reduced to the specified level. If you think about a double-decker bus that’s too high to pass through a tunnel, instead of the tunnel taking the top deck off the bus, it magically squeezes the bus until it’s the right size to fit. So, just as the height of the bus is reduced in our somewhat imaginative metaphor, the loudness, or height of any peaks in the signal is reduced. This is the core of the concept but there are many variations of the technique – some of which will be discussed a little later (see Extensive Compressors).
The terms listed below won’t necessarily apply to every compressor you ever encounter, for example, the stompbox format compressors for basses and guitars just don’t have the space for all the functions, but it gives you a good overview of what you might come across.
The ‘Input’ literally controls the level of the signal being ‘put in’. So how loud the signal is as it enters the compressor. It’s a good idea to set this so that the entire dynamic range is covered. If you already know that there’s going to be a peak in the input signal, make sure that the peak matches the input level – so the maximum level entering the compressor, without getting distorted.
The Threshold sets the point at which the compressor starts to work. This happens when the incoming signal reaches a louder level than the threshold point. So, if the threshold is set to 0dB, the compressor will basically do nothing. If it’s set to -3dB, then compressor will start working whenever the incoming signal passes over the -3dB point. Extreme threshold values like -30dB will, in turn, result in extreme compression.
So, the compressor has detected a peak in the signal and wants to squeeze it down. By setting the ratio, you determine how much the signal is pulled back to squeeze the peak down. By using a 1:1 ratio, nothing will happen. By using a 2:1 ratio, the loudness is halved. By using a 4:1 ratio, three quarters of the loudness is removed, and by using a ∞:1 ratio (infinite:1), any dynamic shifts are completely removed. Using an infinite ration has the most extreme effect, while the closer the first number gets to 1, the more subtle the effect.
De tabel bij ratio toont een strakke/haakse ombuiging: “hard-knee”. Dit kantelpunt zou ook wat meer een curve kunnen hebben, in dat geval heet het “soft-knee”. De invloed op de klank is bij soft-knee wat vloeiender. Er zijn geen echte toepassingsregels; gebruik wat het beste klinkt, en beoordeel het per situatie/instrument.
The attack dictates how quickly the loudness is reduced once the compressor starts doing its thing. Setting the attack time so that it enhances the specific ‘energy’ of an instrument or performance is an art.
It might now be no surprise to know the release controls the speed at which the loudness is returned to its original level. It’s the release paramater that’s used to create that ‘pumping’ effect used in electronic dance.
Make-Up Gain, Output
Hopefully, it should already be pretty clear that compression reduces the dynamic range of a signal by eliminating any peaks. In principle, this also means that the resulting ‘compressed’ signal is actually quieter. A gain, make-up gain, or output control is therefore sometimes provided so that the volume can be pushed back up. In the world of software, and sometimes the world of more expensive hardware like the API 2500, you often get offered the convenient service of automatic make-up gain. This continuously ensures that the compressed signal maintains maximum volume. In fact, this has the effect of boosting the volume of ‘quieter’ signals while signals that were loud enough in the first place will pretty much remain the same. Of course, this depends on how you look at it. Not every engineer will be so trusting of automatic make-up gain, or just don’t see it as all that beneficial. As such, it’s not actually included on the control panel of many hardware compressors and tends to stick to the realm of software compression.
It’s also worth mentioning the phenomenon of the limiter. A limiter is basically a variation on the theme of compression where, in theory, a limiter operates on a ratio of ∞:1 (infinity:1). In practice, engineers already see a compressor-limiter with a ratio of 10:1 as a limiter. A limiter basically takes on the role as problem-solver and is a little less creative than compression. Radio stations use them to maximise a signal and consumer electronics use limiters to ensure that the volume can never get too loud and cause hearing damage. In the world of software, you might also come across the term ‘brickwall-limiter’. This is a simple method used to raise the loudness of a signal and where any peaks are literally cut-off, or clipped – hence the term, ‘clipping’. This can be really subtle, but when more extreme, it’ll be heard as a distorted signal. Instruments like drums are an extreme example, while the effect will be barely noticeable with more subtle sounds like a piano played softly.
Just as equalisation and reverb are largely creative effects, compression is also used to solve certain problems, some of which have been outlined below.
Since vocalists can sometimes sway or move their heads from side to side as they perform a take, this can create small volume differences within the recording – not something an engineer wants! By connecting a compressor to the microphone (like a microphone preamp), this makes sure that the recording retains a consistent level. An exception to this is something like a big band playing on stage with a trumpet section. To put on a good show, the players will often move the horn of the trumpet from side to side, no matter where the microphone is pointed. Here, adding a compressor will make little difference since, as soon as the horn is pointed off to one side of the mic, the signal loses around 20dBs – which a compressor simply can’t compensate for. To solve this very particular problem, a clip-on microphone is the best solution.
Back in the day, when tapes were still produced on a massive scale, audio equipment was generally less clean and a little noise was normal. While the noise was always there, the louder the audio signal, the smaller the signal-to-noise ratio became. So, what you wanted was the music (or ‘sound’ in the broadest sense of the word) to be recorded to the tape at as high a volume level as possible. This is totally possible up to a certain point, but what compression does here is make sure that the volume level is consistent.
Radio and Constant Loudness
Put your hands up if you want to listen to classical music in the car! Now, put your hands up if you give up two minutes in! The problem with classical music is its enormous dynamic range. One part is incredibly loud while other parts are so quiet that you might forget that the radio is even switched on. Unfortunately, compression can’t actually help much with this, but it actually explains why pop music works so well on car radios. Pop music is much more consistent when it comes to volume since it’s been mercilessly squeezed through a few compressors by the producer. Then, the radio station puts it through yet another compressor. Because different pieces of music will vary slightly in terms of volume and radio stations need to broadcast at a consistent volume, compression is essential. While we’re on the subject of radio broadcasting, it’s worth mentioning ‘voice-ducking’: an extreme compression setting that lowers the volume of any music being played back, as soon as the DJ starts to speak.
In essense, a compressor regulates the loudness of an incoming signal and makes adjustments when the level reaches a specific point. But this function is actually very basic in terms of what a lot of compressors can actually do. As such, here’s an overview of some of the other functions offered by more extensive compressors:
With a basic compressor, the source audio signal acts as the trigger for the compression. So, if a peak is detected, the compression is activated. With compressors that support side-chaining, it’s actually possible to use a different signal to trigger the compression. In other words: if the master signal is an extremely quiet string-quarter and you use a loud kick drum as the side-chain signal, then any compression of the string-quartet will be triggered by the pulsing of the kick. Ok, so a string-quartet is a fairly extreme example, but when making pretty much any EDM, this technique is used on a daily basis. Side-chaining offers a mass of creative potential and is definitely a plus-point for any compressor.
So, what if you wanted to add compression to an entire, mixed track so that all of the instruments act as the trigger for the compression? Here, you run the risk of the kick drum undermining every possible ‘trigger-moment’. You could use mix busses to send the kick to its own bus while all the other insturments are sent to a separate one, so that the instruments can be compressed without any interruption from the kick. Or, there are compressors complete with a built-in low-cut filter that’s designed to solve precisely this problem. The Empirical Labs EL8-X Distressor, for example, offers this useful function, along with compressors that usually cost a little more.
A multiband-compressor splits the audio signal over multiple frequency ranges – four is the usual amount. The individual frequency ranges can be adjusted and can also overlap. Here, you get four audio signals and a dedicated compressor per signal – each with their own settings. A software version of multiband compression is easier to come by and iZotope Ozone is a great example of one.
More Creative Compression
A compressor is not just a problem-solver, and there are a lot of great creative things you can do with a little compression and there are no set rules in terms of what you can do. But, here are a few little things that producers before you have already stumbled across:
One of the most well-known effects that can be created using compression is pumping. This has an incredibly effective, uplifting feel to it and really makes the bass move. Is this a new thing? Not exactly. Back in the ‘70s and ‘80s, people would use the swell pedal of an electronic organ to control the loudness, which acted as a kind of manual master-compression. The most logical choice for any electronic dance music producer is a compressor with a side-chaining function. The kick drum (on its own track) can then be used to act as a trigger for the compression of the rest of the arrangement (on a different track).
The Power of Softness
Another great use of compression is its ability to make soft-sounding instruments a bit louder. When an instrument is played softly, it can sound a lot warmer and actually more impacting than if every note is struck with full brutal force. This is actually a great tip from none other than soundtrack composer, Hans Zimmer! With a compressor with a good make-up gain, softer-sounding instruments can be made louder while compromising nothing on their warmth and actually adding to the vibe of an entire arrangement. Naturally, this can depend on personal preferences, but a lot of people generally find that when compression is used, music doesn’t just sound louder, but fatter, rounder, and more present.
In general, the pre-loaded kick drum samples of a keyboard or synth already sound tight and snappy. You know that it sounds ‘snappy’ when you hear a little ‘tick’ at the beginning of each hit. This is the tick that sits on top of the mix, giving the beat plenty of presence. Since other instruments build the depth of the track, the kick drum doesn’t have to be the master of the universe here. If you’ve ever had the pleasure of playing a real kick drum, you’ll have noticed that it sounds completely different; almost ‘flabby’. So how do you get it to snap? Cue the compressor! Using compression, you can emphasise the start of the kick with a fast attack, a slow release, and a hefty ratio. Because the compressor is working fast, the basic sound of the kick is shortened. The output/gain can then be turned up to make the original extra loud and the result is a kick that’s been transformed from an ‘umph’ into a snapping ‘pff’ sound. You can treat the snare in the same way but with toms, it’s a little less effective. Cymbals are better left alone as far as compression is concerned. However, if you’ve recorded an acoustic kit and used overhead microphones to capture the cymbals, then it’s worth using a little compression to tighten things up. If you tend to use tightened synthesized kicks and snares, then they can literally sound synthetic. A snappy sound is not so hard to create when working with synthsized sounds while you can assume that bands will pull their kick and snare through compression before they’re happy with it. A lot of beginner producers have no awareness of this and are often forced to learn the hard way.
A Chain of Compressors
This is another tip we’ve pulled from the producer’s bag of tricks! You’ve probably noticed that a lot of commercial music sounds pretty loud and, now that you have an idea of how compression works, you might also ask why you don’t actually hear that ‘pumping’ effect more. What producers tend to do is chain a number of compressors in series. The output of compressor #1 is connected to compressor #2, and the output from compressor #2 is then hooked up to compressor #3 and so on. When the parameters of each compressor are set to have a subtle effect, this technique can achieve remarkably good results. What you get is something like what we described earlier as a ‘soft-knee’. But where the curve of a soft-knee is quite limited, a chain of compressors yields a much larger curve. When mastering, this method is used a lot but since mastering is an art in its own right, we’re not going to dive too deeply into it here.
There’s another interesting use for compression, but it can seem a little too out-there for most. But, we know you want to get creative so we’ll just lay it out. Basically, anyone working with making short samples for smaller samplers (like retro computers and samplers) definitely benefit from a little compression magic. Instruments like the piano, guitar, vibraphone, or harp have a natural decay (so, the way a note gradually gets quieter and dies out). However, samplers that aren’t gifted with massive memories much prefer samples that can be looped. But looping a sample that’s ‘dying’ at the end and has much less energy at the beginning of the loop doesn’t really work out. Looping a small piece of the sample is one solution but this means removing something from the sound. A sample that last a couple of seconds is therefore better, and to flatten the dynamics (or energy) of the sample, compression can be used. This makes looping the sample much easier. Of course, you do lose that ‘natural death’ of the sound, but a regular envelope can solve this. So, if you grew up in the ‘80s or ‘90s, or you’re busy making your own samples, then you can maybe make good use of these tips – which is why they’re here.
If you send an audio signal through a compressor, it’s like someone is constantly sitting with a watchful finger on the fader. Loud parts become quieter, reducing the dynamic range. Simple, right? But then why are there so many different kinds of compressors available? And why are old models from the ‘50s, ‘60s, and ‘70s so popular?
The character of a compressor is, for the most part, dependent on the way it detects the strength of the signal. Further, you’ve got factors like the attack, release, and ratio strengths of different models. One model will have an entirely linear way of functioning and will maintain the release-time and ratio you’ve set, no matter how much and for how long the source signal has popped over the threshold, while the bahaviour of other models can depend entirely on the source signal. Every compressor will come with an amplifier on board to take care of the make-up gain since the device naturally weakens the signal and needs to strengthen it again before sending it out. It’s also essential to know that valve compression sound very different to transistor compression – so this also plays a big role in the overall sound character of a compressor.
We’ll kick of with VCA compressors, since the layout and sound is pretty similar to most of the standard compressors you get with any DAW. VCA stand for Voltage Controlled Amplifier. A VCA chip works really quickly and can be set with minute precision. All the parameters, like the ratio, attack, and release are offered by any average VCA compressor. These compressors are outstanding problem solvers and are often used to pull back particularly sharp peaks. Since they do this in such a reliable and transparent way, a lot of producers love using them for the master bus. Renowned VCA compressors include the DBX 160 and SSL G-Series bus compressor but these things are designed to perform like studio workhorses, so they’re not necessarily that creative. If you want to push things further and you’re looking for that sparkle and warmth, then it’s best to read on.
Here, we’re going to take the liberty of talking about just one particular compressor: the UREI 1176 (which you can see in the image above). Many clones of the studio-based hero have been made in both hardware and software plug-in form, so if you’re tempted, you should have no trouble finding one. FET stands for Field Effect Transistor, which refers to a type of transistor that has the same character of sound as a vacuum tube. While the operation is comparable to that of a VCA compressor, FET compression packs way more colour. This is due to the behaviour of the harmonic distortion and a relentless attack and release. The 1176, therefore, is like the copper who would arrest you for public indecency while you’re trying to report that someone has nicked your clothes. FET compressors are perfect for treating vocals and drums to give them more attitude. It’s also the perfect compressor for just smashing things with. The famous ‘all-buttons-in’ mode is a popular trick for transforming drum room mics to the point where the sound becomes unrecognisable.
Opto compressors take a little bit of a detour. The incoming voltage (generated by the incoming signal) is used to light a bulb and this light is then collected by a light-sensitive cell. The strength of the detected light then serves as an indication for when the opto compression needs to launch into action and how much the signal needs attenuating. However, since the bulb will never burn at full brightness and actually never goes out completely, opto compressors are actually quite slow. Fortunately, this actually does wonders for vocals and bass guitars. You can’t really go wrong using an opto compressor. Most models will only offer control over the threshold and make-up gain, the ratio is dependent on the loudness of the signal as it enters, and the release-time depends on how long the threshold is exceeded for. As such, by using extreme parameters, an opto compressor has quite a harmonic, musical effect and will never ‘pump’. Opto compressors also use valves for the make-up gain, giving the vocal or instrument track some nice added saturation. The LA-3A, on the other hand, has a transistor rather than a valve amplifier fitted, so sounds cleaner and is generally much faster.
The delta-mu compressor is a valve compressor, meaning that the compression is created by a variable resistor in a tube circuit. This kind of compression is known for its ‘program dependent’ operation and as such, always offers control over the ratio because the amount of compression depends on the strength of the incoming signal. Delta-mu, also known as variable-mu compressors, react quickly to peaks, but not quite so relentlessly as VCA and FET compressors. Rather than clipping the top off a peak, a delta-mu rounds it off. Like FET compression, they’re much loved for their effect on vocals and bass guitars and you’ll often find one in any good mastering studio. This type of compression is able to apply a considerable amount of gain reduction without rendering the original dynamics unrecognisable. It also has a gift for enriching audio with a nice dose of saturation. One of the first compressors ever made was the Fairchild 670 (seen in the image above), and it’s still seen as the holy grail of all compressors, and it’s because of this that the price of a Fairchild is pretty high. More modern delta-mu compressors include the SPL Iron and the Manley Labs Vari-Mu compressors. The Vari-Mu has become such a household name in the production world that the name, ‘Vari-Mu’ is now synonymous with this particular flavour of compression.
What’s your favourite compressor ever? Got any great compression-based tips and tricks? Let us know in the comments!