Many an engineer has found compression to be one of the most mysterious processes in audio. You can turn the knobs on a compressor until it does something that might work in the track, but imagine how much better your mix could sound if you knew precisely the reasoning behind every control, then the best way to set them up.
Also available in this series:
- Compression Tricks Of The Pros - Part 1
- Compression Tricks Of The Pros - Part 2
- Compression Tricks Of The Pros - Part 3
- Compression Tricks Of The Pros - Part 4
In this new series we’re going to look at compressors, limiters, gates and de-essers, then make them work better than you ever thought they could. In Part 1 of this series we’ll discuss the various compressor controls, and the best way to set one up.
So what is a compressor exactly? When you get right down to it, a compressor is nothing more than an automated level control that uses the input signal to determine the output level. The better question really is, “What exactly are we trying to accomplish by using a compressor?”.
First and foremost, a compressor is used to control the dynamics of an audio signal. This means keeping the level of the sound even by lifting the volume of the soft passages and lowering the volume of the loud ones so that there’s less of a difference between them. A compressor can also be used to color the sound as well, but that’s usually a byproduct of the compressor circuitry or programming.
What gets confusing is that there are different categories of compressors that revolve around the way they accomplish the task, so not all compressors sound the same as a result. Regardless of how they sound, all have roughly the same parameter controls and are operated the same way.
Most hardware or software compressors have the same basic controls, although certain controls may have different names from manufacturer to manufacturer. Let’s look at some of typical parameters that you’ll find.
The Threshold control determines the signal level where the compressor starts to work (see Figure 1). Below this threshold point, no compression occurs. For instance, many compressors are calibrated in dB, so a setting of -5dB means that when the level reaches -5dB on the input meter, the compression begins to kick in.
The Ratio parameter controls how much the output level of the compressor will increase compared to the level being fed to the input (see Figure 1). For instance, if the compression ratio is set at 4:1 (four to one), that means for every 4dB of level that goes into the compressor and is above the threshold level, only 1dB will come out. If a compression ratio is set at 12:1, then for every 12dB that goes into the unit, only 1dB will come out of the output.
On some compressors like the famous UREI LA-2A (see Figure 2) and LA-3, the ratio control is fixed so you can’t vary it, but that’s a big reason for the way they sound. That said, on most compressors the Ratio parameter is variable from 1:1 (where there’s no compression) to as much as 100:1, at which point it becomes a limiter (a subject that we’ll cover soon).
Attack And Release
Most, but not all, compressors have Attack and Release controls, which determine how fast or slow the compressor reacts to the beginning (the attack) and end (the release) of the signal envelope. Some compressors (again like the famous UREI LA-2A or the dbx 160A (see Figure 3) have a fixed attack and release that can’t be altered, and this helps to give the compressor its distinctive sound. Other compressors like the famous Fairchild 670 (see Figure 4) have selectable attack and release parameters.
Many compressors have an Auto mode that automatically sets the attack and release according to the dynamics of the signal. Although Auto works relatively well, it still doesn’t allow you to dial in the precise settings required by certain program material.
The Attack and Release controls are the key to proper compressor setup, but many engineers aren’t really sure exactly how that’s done (we’ll go over that in a bit). You can still get reasonable results by keeping these controls set to somewhere around their mid position, but learning just how they work provides more consistent and professional results.
Gain, Make-up Gain, Output
One of the byproducts of compression is that the level is decreased when the signal is compressed, so there needs to be a way to boost the signal back to where it was before. Depending upon the make and model of compressor, this control is called either Gain, Make-Up Gain or Output. It’s frequently used to boost the output in a variety of situations (like mastering for one), since most compressors have a lot of gain available if you need it.
The Gain Reduction Meter
The gain reduction meter indicates how much compression is occurring at any given moment. On most devices this is shown on a VU or peak meter that reads backwards, which is sometimes hard for a beginning engineer to get a handle on. In other words, the meter reads zero when there’s no compression and usually travels to the left or down into the minus range to show the amount of compression. As an example, a meter that reads -4dB indicates that there’s 4dB of compression occurring at that time (see Figure 5)
The Side Chain
Many compressors have an additional input and output called a side chain, which is used for connecting other signal processors to it’s internal compression circuitry. That means that the connected processor only gets sent a signal when the compressor exceeds the threshold and begins to compress.
The side chain is often connected to an EQ to make a de-esser, which will soften the loud "SSS" sounds from a vocalist when they exceed the compressors threshold (we’ll cover this in depth a bit later). It’s possible to connect a delay, reverb or any other type of processor to the side chain for some unusual, program level-dependent effects. Since a side chain isn’t needed for normal compressor operations, many hardware manufacturers choose not to include it on their units since it adds to the cost, but you see it on more and more compressor plugins, even when the original hardware unit didn’t have it available.
Most compressors have a Bypass control that allows you to hear the signal without any gain reduction taking place, then compare it to the compressed audio. This is important because it helps you hear just how much the compressor is controlling or changing the sound. It also makes it easy to set the Output control so the compressed signal is the same level as the uncompressed signal, if that’s what you want.
Setting the Compressor
The most important part of setting up a compressor properly is setting the timing of the attack and release, so here are a few steps to get you there. One of the easiest ways to do that is to start with the snare drum and use that setting as a template for other instruments, assuming that you’re in the studio mixing a song with a more or less constant tempo. Remember, the idea is to make the compressor breathe in time with the song.
- Start with the attack time set as slow as possible, and release time set as fast as possible on the compressor.
- Turn the attack faster until the instrument (in this case, the snare) begins to sound dull (this happens because you’re compressing the attack portion of the sound envelope). Stop increasing the attack time at this point and even back it off a little so the sound stays crisp.
- Adjust the release time so that after the snare hits, the volume goes back to at least 90 percent of the normal level by the next snare beat.
- The more wild the peaks, the higher the ratio control must be set, so increase it until the sound of the snare hits (or whatever instrument or vocal you’re working on) are pretty much the same level.
- Add the rest of the mix back in and listen. Make slight adjustments to the attack and release times and ratio control as needed.
- Use that as a basic setting for the other instruments and vocals in the song. Remember, the wilder the peaks, the higher the ratio needs to be set.
How Much Compression Do I Need?
The amount of compression you use depends upon how uneven the volume of the part is, but remember that the more compression you use, the more likely you’ll hear it working. As a general rule, compression of 6dB or less is used more for controlling dynamics than for imparting any sonic quality, but it’s also common to see as much as 15 or even 20dB used for electric guitars, room mics, drums, and even vocals, depending upon the situation. As with most recording, the amount of compression depends on the song, the arrangement, the player, the room, the instrument or vocalist, or the sound you’re looking for.
In the Part 2 we’ll look at some specific examples of using compression, as well as some times when radical compression can come in handy. Some of the above material comes from my book Mixing And Mastering With IK Multimedia’s T-RackS, and my upcoming book The Audio Mixer’s Bootcamp. You can read excerpts at bobbyowsinski.com.
You can find more music marketing tips on my Music 3.0 music industry blog. For music and production, check out my Big Picture production blog. To read some additional book excerpts as well as some from my other books, go to bobbyowsinski.com. You can also follow me on Twitter for daily blog updates.