Now I have a confession to make, compression is one of those things that I put off learning about for a long time. You may also be one of those people, finally plucking up the courage to read about the subject, if so, welcome! you’re certainly not alone.
I knew it was something I ‘should’ be using. But it was something I would just throw in at the end of the mixing process, and to be honest, I didn’t know if it was helping or hindering the sound of the track. And yes I would occasionally get some good sounding tracks, but I had no idea how, so the next time I came to record I had the same pain all over again.
It turns out that it is not actually that difficult to understand. So I wanted to write a ‘complete guide’ to compression which will try and explain firstly what they are, then how they work and finally provide a few starting points and tips.
But to understand compression you need to know about dynamic range. So…
What Is The Dynamic Range?
You may hear a compressor also being called a dynamic range compressor, but what is the dynamic range?
The dynamic range of a track is the range between the noise floor (background noise) to the loudest possible signal created without the sound distorting.
So throughout a track, this could vary quite considerably between quiet sections and loud sections. This is not really that desirable and can be particularly noticeable on certain tracks such as vocals, or for instruments such as bass where the volume of individual notes tends to vary quite a bit throughout a track.
The image below shows a vocal track with quite a varying dynamic range.
A track like this will therefore probably benefit from some compression, i.e getting a more consistent range of sound throughout the track.
What Does Audio Compression Do?
So in very basic terms, imagine compression as if someone has control of the gain knob and can turn it up whenever the sound gets too quiet or down whenever it gets too loud. Except instead of what would be a frantic panic, it will be done with extreme precision. Not only that, but the speed at which it is done can be altered and loads of other things too.
As with most things these days there are hardware and software compressors. For ease of demonstration, I will be using screenshots from a simple software compressor, in this case it is the one that comes built into Ableton Live. Most compressors in most DAWs will look similar to this and will have most of the same knobs and settings.
Below is a screenshot of what a compressor will look like in most DAWs with a sort of graph and then a meter for ‘threshold’ and knobs for ration, attack and release.
So here is a quick summary of the different settings you are likely to find on most compressors and what they all mean.
As a compressor is used to even out the sound it does this by automatically turning the loudest parts down (by the way, I’m talking about downward compression here, there is also upward compression which I will explain later).
Now you don’t want the compressor turning every note down, as that wouldn’t make any difference and therefore we set a ‘threshold’. When a signal rises past the ‘threshold’ value the compressor kicks in. Below the threshold value (usually measured in decibels (dB)) no compression occurs.
In the below example the threshold has been set to -14.5dB, so the compressor will cut in whenever the signal reaches -14.5dB. The yellow dot on the graph represents the threshold value.
As you can hopefully see in the GIF below, the yellow dot going up and down is the volume of the signal. Every time the yellow dot passes the larger yellow circle (which represents the threshold) the compressor kicks in as can be seen by the bar that says ‘GR’ which stands for gain reduction.
The graph below may help to emphasize the point further. As you can see every time the track (grey bottom graph) goes over the threshold (orange line) the gain is reduced at that point (top graph, yellow line).
For most situations in music, gain is linear. So any increase in level at the input stage is identical to the increase you will then get at the output stage. So increase your input level by 2db and you will get a 2db increase in your output level. A 1:1 ratio.
A compressor can change this ratio (but only in the region of the dynamic range that is above the threshold. So if you have a 4:1 compression ratio a 4db input level will result in a 1db increase at the output. Below the compression threshold the ratio will still remain at 1:1 so it depends how loud the sound is coming in and where you have the threshold set.
Compression ratios can be set at 5:1, 10:1 all the way up to infinity. As the ratio rises you need more sound at the input to get even a quiet sound at the output.
If you use synths regularly you will be familiar with the terms attack and decay/ release. Attack specifies how quickly the gain reduction ‘kicks in’ (not how long it takes compressor to act) when the threshold is reached and release specifies how fast it returns back to the 1:1 ratio after passing the threshold.
The attack is usually measured in milliseconds, giving you an idea of how quick it is (we aren’t talking in seconds here).
When to use fast attack?
I use a fast attack most of the time, as I want the compressor to kick in immediately. A fast attack is good if you want to get rid of any sharp transients which often occur at the start of each note. If you are unfamiliar with the word transient, think of a drummer hitting a crash cymbal, the sharp transient is the initial spike in volume after the cymbal is hit before it slowly rings out.
When to use a slow attack?
A slower attack is useful when you want to compress tail end sounds and leave those transient sounds untouched (cutting transients can make tracks lose the punch you desire). Be careful as if it is too long it may compress the next transient and leave you in a bit of a mess. Often on a snare or other drum track, you will want that transient to come through to add punch to the track and so a slower attack time may be useful.
To get away from the downsides of both you can use serial compression where maybe you use a fast attack compressor followed by a slow attack compressor. I explain this later on in the article in the section on bass compression.
Playing with the attack setting is one of the best ways to really hear the effects of compression if you are just starting out using it and want to hear the effects in the extreme.
Decay or Release
The decay or release is how quickly the compressor stops attenuating the sound once the signal level falls below the threshold.
Setting this value very short is called ‘pumping’ causing a very artificial sound.
The compressor works by bringing the loudest parts of the track down and so you will more than likely end up with a quieter track. This is where the makeup gain comes into play. This allows you to bring the whole track back up to a level where it can be heard in the mix once more.
In compression, you may hear the term ‘knee’. There are two types of knee, a soft knee and hard knee. And the ‘hardness’ of the knee basically determines how extreme the transition into compression is. So a hard knee is more dramatic than a soft knee.
On hardware compressors, it is often just one or the other, hard or soft. But these days in our DAWs you are likely to have a knob and so it is not quite so black and white.
Below is an example of a harder knee.
A softer knee (like below) makes the transition from no compression to compression more subtle. This is most appropriate on tracks with a larger dynamic range such as vocals or acoustic guitar.
How Do I Know Which Compression Settings To Use?
It’s all well and good knowing what all the terminology means but how does that actually apply to creating music?
Here is the basic technique I use when inserting a compressor into the signal chain.
- Move your threshold up and down until the loudest parts of the signal are triggering the compressor.
- Set the ratio – this will depend on what you are recording, I tend to use 2:1 for most things, but try 4:1 and 6:1 to see what sounds right.
- Now adjust the attack and release, again there is no best value, see the sections below for more specifics.
More Advanced Compression Techniques
I’m not going to go into loads of detail here but I wanted to give a brief introduction to some different compression techniques that you may hear about or come across. They are more advanced and can ruin your music if not used correctly so I warn you to approach and dabble with caution.
What Is Multi-Band Compression?
A multi-band compressor allows you to compress a track at a range of different frequencies. So you may have one compressor acting between 0hz and 50hz and then between 50hz and 500hz with another and so on. This allows you to apply different amounts or types of compression to certain frequency ranges of that particular track.
When might multi-band compression be useful?
Tonal inconsistency in vocals is a common place where multi-band compression is used. You may want to compress the track at 3 or 4 different places to really shape the sound you want.
It is also a common way for getting rid of annoying sounds that may creep into your vocals. Doing this using just EQ may get rid of other sounds that you want to keep and so if you just apply compression to the annoying frequencies it will compress them when they are too loud rather than eliminating them all together.
You may also want to use it on a drum bus/ return track to even out your drum sound.
Be warned, multi-band compression tends to get overused by many people, don’t get carried away and throw it on all your tracks automatically. For this reason, I only use it to solve issues in the sound rather than using it every time.
What Is Sidechain Compression?
You may be familiar with sidechains already. In compression a sidechain is basically using other tracks to control the compressor settings for you automatically. So instead of simply setting the threshold and then cutting in every time that is reached, the compressor instead uses another track as the cue for when it should be activated.
When might sidechain compression be useful?
This is another way to bring out certain elements of a track on top of techniques such as altering EQ. So, for example, you could use a drum track to control a compressor on the bass track.
With the sidechain compressor set on the bass track to be controlled by the kick drum, every time the kick drum sounds, the compressor will be activated on the bass. If the compressor is set to bring the gain of the bass down at that particular point then it means the kick drum stands out more in the mix and doesn’t clash with the bass and give you a muddy mix.
A second situation when it may be useful is to get rid of unwanted sounds on certain microphones. So say you had a drum kit mic’ed up with overhead mics and close mics those microphones are going to pick up sounds from all the drums. Many drummers only want the snare sound to come through the close snare mic and not to get picked up in the overhead mics. A sidechain compressor may be able to help in this situation. Set a sidechain compressor on the overhead mics which will be triggered by the snare microphone and then every time there is a hit on the snare the overhead mics will be compressed at that point and you should just get less snare drum on the overhead mic track.
The trick is to use this setting sparingly, you don’t want to lose a track completely you just want to carve out a bit of space in the mix.
What Is Parallel Compression?
Parallel compression is the process of taking a track, duplicating it (so you have two of the same) and then applying compression to just one of these two tracks. This allows you to blend between a compressed and uncompressed track whilst mixing, in theory giving you the best of both worlds.
When might parallel compression be useful?
The main benefits of parallel compression is to give extra impact and energy to a track without muddying up the mix. One of the most popular uses would be on the drum bus/ return track.
To get the full benefit of this you need to use a fast attack time to get nice consistency in the track accompanied by a boost to the bass and treble (depending on what it is you are trying to bring out in the mix). The uncompressed track will allow the transients through whilst the heavily compressed track will give you more consistency.
Compression Tips and Tricks
In the final section of this article, I am going to give you a few tips and tricks to help you get started with using compression. This includes some short sections on specific techniques and settings for vocals, bass and drums.
These are just some starting points and aren’t meant to be complete guides for each category (that would end up being a book rather than a blog post).
How To Use Compression On Vocals
Quite often you will find that vocals get lost in your mix. Which is not really what we want, after all the vocal is meant to be what stands out, and what people sing along to!
I used to have this problem a lot. Its confession time again…I used to try and solve this by cranking up the entire track and then it would solve the problem in some areas but then in areas where the singing was louder like the start of a chorus there would be a peak and the sound would become distorted. I would then go through manually tweaking these peaks. It sounds silly now but I’ve made these mistakes so you don’t have to!
Certain sounds, vowel sounds for instance, just tend to naturally come out louder than some softer letters. So this is the dynamic range I was talking about earlier. You want to even this out but without losing the expression and tone within the track.
Well, the compressor helps us out in this situation.
Now as with anything in music just giving you specific numbers and settings isn’t really possible as every track is different but I can at least give you some starting points.
So start by setting the threshold just below the loudest part of the vocal track. Move the knob up and down until you see a gain reduction of around -6db.
For the ratio, in general, 2:1 is the best ratio for vocals but you can test out 3:1 or 4:1 (it depends on the style), anything other than this and it doesn’t really sound natural.
Have a listen now. Are you losing clarity in some of the louder more detailed notes? In this case, try increasing the attack value to allow some more of those peaks to sneak through before the compressor is allowed to kick in. In terms of release in line with the natural release of the vocal, again there is no exact value just see what sound best.
As mentioned above once you have added these settings you will probably need to use some makeup gain to bring the vocal level back up so you can hear it in the mix.
In terms of the knee. A softer knee is usually better on vocals as it still gives you a bit of dynamic range as the compressor more gradually comes in after the threshold is reached rather than straight away.
How To Use Compression On Bass
Compression on bass is common in mixing to get your bass to sound a ‘fuller’ and ‘fatter’, i,e standing out a bit more in the mix. If you play bass you will know it is naturally very dynamic and you will pluck certain notes and sections louder even if this is not always intended.
Most modern styles of music use a driving bass that is easy to pick out by ear and this is what most people desire (rock, hip-hop, dance music etc.). To get this compression allows you to even out the dynamic range and then increase the overall volume of the track without having distorted or unwanted louder sections. Consistency is key.
As I mentioned with vocals before I can’t give you a starting value for threshold or gain because that will depend on your particular track, however, I can give you a few starting points.
- For bass, a compression ratio of 4:1 is usually a good start but you can go much higher depending on the dynamic range on your bass track.
- Start with no attack and slowly increase it until you notice that the bass starts to lose a bit of character and life. If it does then you are losing some of those transients and the attack is too high, always stay on the lower end of the attack range if you are uncertain.
- The release time is important for bass as it is what will give you sustain if that is what you desire. Say you have long steady bass notes that ring out, which is common in rock music, you don’t want too short a release on the compressor as you will lose that sustain. A good starting point is between 100ms and 150ms for bass and then work from there.
For bass, it is also worth considering using serial compression.
What is serial compression?
Serial compression is the act of using more than one compressor to compress the track at more than one point. As the name suggests they act in series so one compressor after the next and so on.
But why? Well on a track with a big dynamic range you may put a faster attack higher ratio compressor first to tame any big peaks and give you consistency and control. But this can leave you with an over-compressed, lifeless sound. So by taking this sound which has been compressed and running it through a second slower attack, lower ratio compressor. This gives us the best of both worlds, consistency but without losing life in the track.
So with this theory in mind, a general rule of thumb put a fast compressor before slower compressors and higher ratio compressors before lower ratio ones.
How To Use Compression On Drums
Drum compression can be very complex and can be used in many different ways depending on the type of drum or other factors such as are you using live drums or programmed.
I’ve already spoken earlier in the article about how multi-band , sidechain and parallel compression can be used to enhance your drum tracks.
In general though try not to overcompress your drums. When you are starting out with compression use it with caution and sparingly. I would suggest just using compression on the drum bus (i.e a return track with all the drums going through) in order to get rid of some of the more extreme transients on the snare and cymbals. Use a ratio of around 4:1 to start with a fairly fast attack time to catch the transients.
When you get more confident start trying the more advanced techniques.
Other Quick Tips When Using Compression
Apply the compressor in context
Remember to always mix in context. You want to be applying the compressor whilst your other tracks are playing otherwise you will not know how the tracks sit together. It is all well and good spending hours tweaking the compression on a track and getting it to sound great on its own. But if it doesn’t work in contrast with the other tracks, you have just wasted a lot of time.
Volume balance before you compress
Don’t start compressing until you’ve sorted your volume balancing first. It may be a case that simply turning tracks up or down at the start can save you a lot of headache at the end.
Don’t be afraid to use presets when starting out
Some people will say that using presets is not very professional, but if you are new to compression it is a good way to learn. Most DAWs will have compression settings built in such as ‘kick compressor’ etc. If you add one of these in it will give you a starting point and you can tweak from there.
Upward compression may be handy aswell as downward compression.
Up until this point I have been focussing on downward compression, but there may be times when upward compression is more suitable. Upward compression also helps even out the volume of the track by decreasing the dynamic range, but instead of pulling down the loud parts when the signal goes above a threshold, the signal is boosted when it is below a certain threshold.
An expander might be better than a compressor in some situations
It may not surprise you that expansion is the opposite of compression. Expansion may be a quick way to pull out certain elements in the mix, such as to give more punch to a particular drum track.
If your sound is a square wave sound it doesn’t need compressing
A square wave sound is basically already compressed.
Hardware vs software compressor
This is so often the question these days as most things move onto the computer and the amount of physical equipment we have in our home music studios continues to decrease.
But is there ever still a reason to go for a hardware compressor?
Well, some hardware compressors are pretty pricey so there must be something to them otherwise everyone would use the one that comes free with the DAW right? Well, I guess it is down to personal taste as with a lot of things in music production. The same reason why you might buy an analog synth over a digital synth for example, or use guitar pedals instead of DAW effects.
Some people just like the physical nature of being able to twist and turn the knobs until it sounds just right.
Another advantage when recording is you can prevent clipping. By having the compressor in the signal chain before the interface you know that no matter how loud you sing or play the compressor will ensure that it doesn’t distort and clip and ruin that perfect take.
But in all honesty for a beginner or for most of us in a home studio setting the software compressors should do the job fine. There are much more important things you can blow your hard earned cash on!