In part one we looked at some easy to use and forgiving compressors. This time we’ll be looking at the controls you’ll find on more typical compressors, which will give you more control to manipulate and shape the sound. When you look at the interface of most compressors, whether they’re plug-ins or hardware units, it’s likely that you’ll see the controls below:
Threshold This is the level above which the sound will be compressed. Sounds that do not reach the threshold level will not be compressed; therefore, if you wanted to compress just the loudest peaks of the audio, you would set a high threshold, whilst a low threshold would lead to heavier compression affect.
Knee When the sound reaches the threshold, how abruptly will the compressor kick in? That’s what the knee control decides. With a hard-knee setting, the compressor will start to work right when the sound reaches the threshold. Soft-knee, as the name suggests, is a gentler, smoother effect, which makes the compressor start to work a bit before the threshold is reached. In short, if you want a harsher, more pronounced compression effect, go for a hard-knee setting; if you want something more subtle and transparent, try soft-knee.
Ratio For sounds that are loud enough to pass over the threshold, the ratio control decides by how much they are compressed. Imagine your sound is a Scooby Doo-style tall sandwich: the harder you press on the top of the sandwich, the more the contents are squished together. Similarly, the higher the ratio, the more the sounds over the threshold will be compressed. A ratio of 4:1, for example, means that the part of the sound passing over the threshold will be compressed to a quarter of its regular volume. Therefore, a high ratio combined with a low threshold will give a heavy compression effect.
If you set the ratio to a really high value, hardly any signal will pass over the threshold. For this reason, a compressor set to a ratio of 10:1 or more is often regarded as a limiter. A compressor that has a ratio of infinity: 1 is called a brickwall limiter, because it limits the signal so that no signal at all will be allowed above the threshold. Some compressors, such as the Sonalksis SV-315 mk2, have a built-in limiter in addition to the compressor. This allows you to use the compressor to shape the sound you want, whilst leaving the limiter to catch any peaks that might otherwise sneak through. We’ll look at using limiters in part five.
Attack So, if you set your compressor up with the ratio set to maximum, would that keep the signal entirely beneath the threshold? Only if the compressor reacts very quickly! The attack setting determines how fast the compressor starts to compress after a signal passes above the threshold. Sometimes – as with brickwall limiting – you do want the signal to be entirely compressed right from the very start. Other times, you might want a slower attack in order to let the initial part of the sound through uncompressed. One popular example is when compressing snare drums: by varying the attack time you can accentuate the sound of the stick hitting the drum head, and get the ‘thwack’ sound that is instantly recognizable.
Often a compressed sound will appear more natural if a slower attack is used. In mastering, for example, the goal is usually to compress the music without the listener noticing the effect too much, and in this case a popular approach is to set a long attack time along with an even longer release time.
Release How long do you want the compression effect to last for? The release setting decides how many seconds or milliseconds it will take for the compressor to allow the volume to return to its original level. Let’s go back to our Scooby Doo-style tall sandwich: if we press down hard on it once, the bread and filling will slowly rise up again after we have released our hand. (If the whole Scooby-Doo thing isn’t working for you, or you had a dismal, cartoon-free youth, just imagine pressing a damp sponge instead.) The time it takes for the sandwich to rise up again is like the release time of the compressor. The faster the release time, the faster the volume rises again after having been compressed.
By varying the release time you can achieve a wide variety of compression effects. For example, as with the long attack setting, a long release time is often used when mastering, because this allows for a transparent and unobtrusive compression effect. Conversely, very short release times can generate audible distortion and give quite aggressive results – an effect that you may often find desirable on some sources. When used in moderation, however, very short release times can also yield transparent results; many digital limiters such as the Waves L2 allow for very fast release settings.
Inbetween the very fast and very slow release settings are a vista of opportunities for sculpting and shaping your sound. A popular setting for the release control is to set it so that the release coincides with the beat and off-beat of the music. This is easy to achieve with any compressor that has visual metering – for example, either a VU meter, LEDs, or a graphical emulation of either – since you just raise or lower the release control until the VU needle or LEDs are moving in time to the music. If you then lower the threshold and increase the ratio, you’ll hear an exaggerated version of the pumping and breathing effect that is so common when referring to compression.
Gain I’ve talked about the ability of compression to make your sound louder and punchier, but clearly if we do nothing except reduce the peaks in volume as I have described then the sound is actually going to be quieter, not louder! That’s why compressors have a make-up gain control, which allows you to raise the overall volume after compression. Many compressors have an auto-gain setting, which handily raises the gain for you automatically without you having to do anything.
If you are setting the make-up gain manually, I find that it’s good practice to set it so that the apparent volume of individual tracks is about the same when compressed as when the compressor is bypassed. That way you can compare the processed and unprocessed versions fairly, and hear exactly what audible difference the compressor is making without being tricked by the increase or decrease in volume (psychologically, louder almost always sounds better).
All the controls together So, how do we combine all the controls together to get the right compression effect for our sounds? The answer is that this is going to vary considerably depending on the source material, and which is why the preset settings found with most compressor plug-ins rarely yield perfect results. The only solution is to use your ears and to experiment. Nevertheless, some general rules still hold, as you will probably have been able to deduce from the descriptions above. So if you want the heaviest compression effect possible, for example, you would set a low threshold, hard knee, high ratio, and a fast release. On the other hand, for something really very subtle, you might try a high threshold, soft-knee, low ratio, and a medium to slow release, possibly with a slower attack. Or, if you want a heavy compression effect but only on the loudest peaks (say, you just want to catch the loudest snare drum hits, for example) then you might set a high threshold but with a hard knee and high ratio. Try different settings on different source material, and you’ll soon develop your own techniques and preferences for how you like your compression to sound.
Special Cases As we saw in part one, not all compressors have the controls listed above. Some set them automatically behind the scenes and just give you a couple of parameters to adjust to make things simpler. Conversely, some have extra controls to vary the effects on offer. These will vary from one compressor to the next. The Sonalksis SV-315 mk2, for example, has an additional control named ‘crush’, which varies the way the compressor reacts to very fast peaks. It also has a built-in frequency-based side-chain (for more on side-chaining, see part four) which enables only certain frequencies to trigger the compressor. One popular variant on this feature is to have a high-pass filter option built into the compressor (as found on the Drawmer 1968 hardware compressor, for example). When this switch is engaged, only the mid and high frequencies in the sound trigger the compression effect. (It will still compress the whole sound, including all those low frequencies – it’s just that the compression effect will not be triggered by them.) This is a useful addition, because the low frequencies contain most of the energy in a track, and sometimes very low – perhaps even inaudible – frequencies can make the compressor work inefficiently. By allowing the compressor to ignore those frequencies, we can often get a clearer and less muddy-sounding compression effect. Which leads us quite neatly on to…
Multi-band compression If one compressor is great, just imagine how fantastic five or so of them working in tandem is! Well, sometimes this is true. A multi-band compressor splits the source into several (usually three, four, or five) frequency bands and compresses them individually before recombining them and outputting them as a whole again. The chief motivation for this is as I mentioned above, that sometimes compression triggered by certain frequencies can negatively affect the sound of others. By compressing each band separately we are able to minimize these effects and end up with a very transparent compression effect. And because we now have control over the individual frequency bands, it’s also possible to radically reshape the sound using a multi-band compressor rather like you might with an EQ. That is why multi-band compression is sometimes used by mastering engineers to fix a track that has been poorly mixed.
However, before you insert multi-band compressors on every track in your recording, you should realize that a multi-band compressor is overkill for most sound sources, which can be compressed much more simply and successfully with a regular compressor. Most engineers will only use a multi-band compressor when all other more straightforward tools have failed. There is a danger – particularly for the beginner – that you might do more harm than good to your sound with a multi-band compressor. A popular idea is that if you like what a multi-band compressor does to your track, it would be advisable to go back to your project and remix it (using conventional EQ and compression) if time allows. Nevertheless, multi-band compression can be a useful tool to have in your arsenal.
Hopefully this chapter has helped you to understand some of the controls and options available when using compression. In parts 3 and 4 I’ll highlight a couple of techniques you can try when, despite your best efforts, those existing controls and options aren’t giving you quite the sound that you’d hoped for…