by Rick Allen
Compressing your audio is a very important tool for enhancing the results of your production. Used properly it can make the difference between that "national" sound and the "local announcer" sound. Use it the wrong way and people end up with those spots that seem like they were mixed through a pair of speakers stuffed with dirty dish towels and cotton balls. Yet compressors still remain mysterious black boxes for most of us.
Of the top ten questions I get asked while I'm in the studio -- coming in right above, "Did you really pay for that sweater?" -- is, "What's your compression setting when you produce sweepers?" To the disappointment of a few people, the answer isn't much help. The kind and amount of compression I use depends on what sound I'm going for in the mix. It's an honest answer, but not very satisfying. So here in black and white I'm going to go a little further in explaining my approach to audio compression in radio production. I'll do that by taking a basic look at what compression is, how it works, what it does, and the different ways you can use it.
A compressor is really an electronic volume control that automatically adjusts the volume of the signal coming out of it based on the volume of the signal going in. The compressor "turns down the volume" as the input level increases. Just how fast and how much is generally controlled by a combination of four control settings on the unit. Compressors usually have a control for threshold, compression ratio, attack, and release time.
The threshold is the signal level you need to reach for the compressor to go to work. Below this level the unit just passes the signal right through. Above this level the "compression ratio" goes into effect. Compression ratio is the relationship between input volume and output volume. Settings vary from 2:1 on up to infinity:1. (Anything above 10:1 is generally labeled "limiting," but that's another story.) With a ratio of 2:1 the input signal must increase 2dB to move the output level up just 1dB. At 7:1 the input must jump 7dB for that same 1dB increase. Attack time is the time it takes the compressor to react to changes in incoming signal and alter its volume. Generally, a short attack time can help a voice cut through. Release time is the length of time it takes a compressor to return the output signal to equal the input signal. Too long a time setting reduces the compression effect, but too short a setting makes a compressor "pump" in a distracting manner. (Ever hear those larger than life announcer breaths?) You also might find a meter that gives you an indication of gain reduction in units of dB. How much gain reduction changes a sound is interrelated to the ratio. 10dB of reduction at 2:1 may sound fine while 10dB at 10:1 might sound totally muffled. Understanding how all these controls interact should help you to adjust your settings to find the sound you want.
Many engineers discourage production people from using compression in the production studio. Some are purists and feel that compressors should only be used for the purpose they were developed for -- to control dynamic range. Still more just don't want DJs messing with compressors because they've heard what happens to their signal when someone in the production studio has no clue how to use compression properly. These are the production guys that don't listen to the final effect of compression over the air. Always remember there are more stages of compression in your station's audio chain after the production studio. Watch out for the people who think that if a little compression is good, tons has to be great. Hey, compression is like water. You're uncomfortable without any around at all. On the other hand, you can drown in too much of it.
There are no right or wrong settings. If what's coming out over the air is the sound you want, then your settings are on the money. If you're interested, I'll give you some ideas of where I start with my compression settings before I begin to tweak. For voice, I work with a middle compression ratio, or 5:1. Bigger ratios can bring up your signal noise by squashing the dynamic range. I then set a fast attack and release time. Just make sure you don't get too short on the release and start hearing the voice track suck in and out. Finally, I'll adjust the threshold until I get a meter indication of about 3 to 4dB of compression on the peaks. (Watch your overall output level. As you turn up the threshold, you generally lose output volume.) Threshold is the setting that will control the amount of "compressing" you do to the voice.
Compressors won't win the war all by themselves, but they're a great part of the audio arsenal. They can't change a below average voice into super sweeper guy, but, when understood and mastered, compressors can help make any voice sound sonically stronger.