XO Wave: How to Use Compression

Compression can increase the average volume of a dynamic recording.

Compression is one of the most important effects available to audio engineers. It is not uncommon for songs you hear on the radio to go through 4 or 5 different compressors before you hear them: one while recording, one or two while mixing, one while mastering, and another at the radio station. Because of its ubiquity, it's important to understand how compression works and how it can be used.

Compression works by reducing the volume of loud passages and increasing the volume of quiet passages. This can be used to bring out subtle details; help control a vocalist with bad mike technique; limit distortion from loud, transient sounds; and give a track more consistent levels, making it easier to mix. Compression can also increase the average level of a track, effectively making it sound louder. Just as importantly, compression is an integral part of how we hear modern music: today's pop music sound simply would not be what it is without compression.

As you read this tutorial, you may want to use XO Wave's compressor to practice these techniques. We also have specific information on XO Wave's Compressor.

Basic Controls

Although there are many variations on the theme, most compressors offer a few basic controls: threshold, ratio, attack, and release. Don't be fooled into thinking more is better: while more controls can offer you more flexibility, it can also take longer to find the right settings, and there are a number of compressors out there with just two or three controls that sound great.

One way to think of a compressor is as automatic volume adjustment. The compressor works by making the loud passages quieter and the quiet passages louder ("compressing" the dynamic range, so the variation from quietest to loudest is decreased). The threshold controls how loud the signal has to be before its volume is reduced. Thus, setting the threshold high means that less compression will occur, because less of the signal will exceed the threshold to activate the compressor. Conversely, a low threshold means that the compressor will do more, because more of the signal will exceed the threshold.

When the signal goes above the threshold, the gain of the signal is reduced by an amount controlled by the ratio control. For example, with the ratio set at 2:1, an increase in input of 2 dB above the threshold will result in an increase in the output of only 1 dB. Higher ratios yield more compression, but only to signals that pass above the threshold.

The attack and release controls work together with the threshold control to determine when compression should begin and end. If the signal is above the threshold for less time than the attack setting, then little or no compression occurs. Once compression is activated, it stays on for a length of time determined by the release control. So by setting a long (slow) attack, you can let through the first part of a loud sound. This is particularly useful because almost every note played on an instrument contains a little burst of energy, called the "initial transient". By setting the attack long enough to let the initial transients through, you can prevent the compressor from squashing the life out of the recording. So if your compressed tracks sound lifeless, try a longer attack setting. On the other hand, setting the attack too long will prevent the compressor from working.

A least half the time, you will be using compression to reduce the dynamic range of your signal, which means that you will be reducing the difference between the loud sections and the quiet sections. This happens in two steps: first, the volume of each loud section is reduced based on your threshold and ratio settings; and then the volume of the whole track is raised based on the make-up gain settings. The net effect is that the loud parts are still loud and the quiet parts are made louder. Compression, therefore, can increase the average volume of the track without increasing the maximum volume.

How to Tweak

Now that you understand the basics, we'll discuss other and more specific uses in the next section, but let's start with reducing dynamics. Of course, every unit and every track is different, but I generally start by setting the ratio low, around 1.5:1, the threshold high, and both the attack and release somewhere in the middle. I then bring the threshold down until the gain reduction meters (or my ears) tell me that the compressor is reducing the signal. From there, I usually tweak the controls bit by bit until I get what I want, which may be a particular sound, or may be a particular reduction in dynamic range.

Attack and release times should generally correspond to the speed of the instrument. For example, bass tracks should normally use slow attack and release times while drum tracks usually sound best with fast times. When adding compression to a mix, rather than an individual track, I generally recommend fast attack and release times, though other engineers will insist that you use slow attack and release times. Everyone agrees, though, to keep the ratio low: too much compression on a mix is a sure way to wreck things!

Once you set the threshold, ratio, attack, and release settings, you should look at the meters and increase the gain to make up for the gain reduction imposed by the compressor. Typically, you'll set the gain on the compressor to be equal to the gain reduction shown on the meters.

A good thing to do is to listen to other mixes of similar music. Some forms of music, especially acoustic music, sound best without any compression at all. Electronic music, punk, and hip-hop often use huge amounts of compression to help them compete with other songs on the radio, as well as to create sounds that are both high-energy and rich in detail. In the case of these modern forms of music, compression is often the key ingredient that puts life into the sounds. Folk, country and other similar genres typically benefit from moderate compression. Country music often uses large amounts on some instruments, such as drums, and less on others.

Obviously experimentation is crucial, not just for learning but for everyday use, since no two tracks are ever the same. On the other hand, many compressors, such as the one included in XO Wave, have built-in presets that give you a nice jumping-off point. Often, you'll only have to set threshold and make-up gain.

Watch those settings!

Other Uses of Compression

Besides reducing dynamics, compression has a variety of uses, and sometimes a reduction in dynamics may be motivated by something in particular, such as a poor performance or an uneven mix. Some of the tricks of compression involve the use of the compressor's side-chain, which is a special signal path that is only used for controlling the compressor. For example, if you put a high-pass EQ on the side-chain, the compressor would become more sensitive to high-frequency content than low-frequency content, and work harder to reduce the signal whenever the signal had a lot of high-frequency content.

Compensating for a Poor Performance: Compression can often be used to overcome weaknesses of the performer. For example, experienced singers move away from the mike when they sing loudly and towards the mike when they sing softly. Less experienced singers generally don't do this at all or, even worse, do the opposite. Even with an experienced singer, it can be desirable to get that "close-miked" sound even during the loud passages. Other instruments suffer, too: it can take years of experience to play evenly and with carefully controlled dynamics. Sometimes even great musicians will do a great take with bad dynamics. In all these cases, compression is essential because the dynamic range of the performance is wider than it should be and compression can bring it back into the correct range.

De-essing: A compressor forms the basis for another effect called a de-esser. A de-esser is used for reducing the so-called sibilant sounds of vocalists, such as "s", "th", and "f", which often have highly exaggerated high-frequency response in recordings. The best thing is always to try getting rid of as much sibilance as possible by using appropriate mike technique, but since this isn't always possible, most engineers like to have a de-esser handy. Note that you can create your own de-esser by using an EQ with a compressor. See the Compressor description for details on doing this in XO Wave.

Ducking: Ducking is a little different from compression: instead of reducing the gain based on the volume of the signal going through it, a ducker reduces the gain based on another signal. For example, you could use ducking in a podcast to reduce the volume of music when someone is speaking. Note that in XO Wave, Ducking is done with the special Envelope Generator effect instead of a compressor.

Compressing a Whole Mix: Compressing a mix is different from compressing a single track because every instrument gets compressed the same amount. In practice, this often means that the bass instruments cause the whole mix to get over-compressed, which makes it sound like the bass instruments are "punching out" the other instruments. Many compressors, including XO Wave's, offer compensation for this. If not, you can compensate by using the compressor's side-chain feature with an EQ.

Compression As a Special Effect: Sometimes compression just sounds cool and that's all there is to it. A lot of hip-hop samples, for example, are heavily compressed. It often sounds good because tiny details suddenly become prominent, while the loud sections are still under control. Compression can also soften the edges of distortion, and enhance performance nuances.

Compression to Increase Dynamic Range: After all this talk of reducing and controlling dynamic range, it may seem strange that compression can also be used to increase dynamic range, but it can be a great tool for this as well. Often a track, such as a snare drum, may sound great by itself, but may have a hard time competing with other instruments in the mix. To help it cut through, you can add a little compression with a relatively long attack time. This reduces the volume of most of the snare drum sound, but keeps the initial transient loud. This short burst of sound is often all a snare drum or other percussion instrument needs to sound great in the mix. Depending on your gain setting, this technique can do two things: it can increase the volume of the transient, helping the track to stand out, or it can reduce the volume of the rest of the note, which may help un-clutter a dense mix.

Conclusion

When I was first learning to use a compressor, I over-compressed everything, which made my early mixes sound lifeless and dull. Some people don't use enough compression, which make the mix sound inconsistent or "jumpy". Finding the right balance is the key, and the only way to do that is to practice and listen, both to your own mixes and the mixes on your favorite CDs.

-- Bjorn Roche