How Do Compressors Work?

Basic Compressor DesignsVariable MuOpticalFETVCACompressor ControlsThresholdAttackReleaseRatioKneeGain MakeupUsing Your Compressor

I once heard Dave Derr, the inventor of the Empirical Labs Distressor, explain how a compressor works by saying, “You can think of it as a little man inside the box who’s turning a volume knob up and down.” He continued, “When it’s too loud, the little man turns the volume down, and when it’s too quiet, he turns the volume up.” I can’t disagree.

Though the earliest compressors were fairly simple automatic gain control (AGC) devices designed to limit overmodulation and distortion in early amplification and broadcast systems, they’ve come a long way. Today, engineers use compressors as tone-shaping devices to add distortion to a signal and, most often, to make things louder — a long way from the original concept of simply keeping signals from being too loud.

Basic Compressor Designs

Variable Mu

Analog compressors can be categorized based on the circuitry used for gain reduction. The first commercially available AGC devices — Western Electric’s 110A and RCA’s 96A — used a tube designed for variable amplification (or “Mu”) that worked by mixing a DC voltage with the audio at the grid of the tube. Though a number of different Variable Mu tubes were used early on, including the 6L7 and 6BA6, General Electric’s 6386, a remote cutoff dual-triode tube, became the standard compressor tube of the 1950s. The 6386 was used in the Gates Sta-Level, RCA’s BA6A, and in Fairchild’s legendary 660 and 670 compressor limiters. Variable Mu compressors have a soft knee (meaning that the onset of gain reduction is fairly gentle), the attack and release times are fairly slow, the compression ratio is program dependent (the higher the signal level, the greater the compression), and the tube used for gain reduction can have an impact on the sound of the audio. Variable Mu designs are still being made, including the Retro Sta-Level and the Manley Variable Mu.

Add to cartLearn MoreAdd to list

Add to cartLearn MoreAdd to list

Optical

The optical limiter was developed by Jim Lawrence, an electrical engineer who founded the Teletronix company after a stint developing optical sensors for the Jet Propulsion Laboratory. Lawrence’s passion was radio and broadcasting, and he came up with the concept of using an optical attenuator to “level” the incoming audio signal so that it remained at a fairly consistent level. This attenuator Lawrence designed, known as the T4, has an electroluminescent panel and a pair of photo resistors. Part of the incoming audio signal is fed to the T4, and as the input level increases, the luminescent panel gets brighter. Light from the panel falls on the photo resistor, which causes the photo resistor’s impedance to increase. The increased resistance reduces the output gain of the original signal. When the input level decreases, the panel dims, which reduces the impedance of the resistor and raises the output volume. The T4 is both program and frequency dependent: while the attack time averages 10ms (milliseconds), the release time varies greatly with the source material. The initial release time is about 60ms for the first 50% of the release, but the remaining 50% of the release happens over a period of between one and 15 seconds. But it’s a very musical variance. With the first version of Teletronix / Optical Leveling Amplifier, the LA-1 only sold about 100 units, but that allowed the company to create a revised version — the LA-2. The third revision of the Teletronix optical limiter was the iconic LA-2A, which, under the auspices of Universal Audio, is still in production. Other optical compressors are currently made by Tube-Tech, Warm Audio, and Manley.

Add to cartLearn MoreAdd to list

Add to cartLearn MoreAdd to list

FET

By 1966, transistors had started to come into their own — audio equipment designers, including Universal Audio’s Bill Putnam, were taking advantage of the newly designed Field Effect Transistor (FET). Putnam’s Universal Audio 175 compressor was based on a Variable Mu design but was reworked with FETs, and the 1176 was born. When it was introduced, the 1176’s major selling point was that its attack time was amazingly fast — only 20 µS (microseconds or.00002 seconds) at its fastest setting, compared to the leisurely 10ms (0.01 seconds) for the LA-2A’s optical gain reduction circuit. The 1176 also had variable attack and release controls and selectable compression ratios. In addition to the FET being used for gain reduction, the 1176 had a distinctive sound due to the Class A output circuit as well as the input and output transformers. While the 1176 wasn’t transparent by any stretch, it had a popular sound — one that remains popular to this day. Universal Audio continues to make an 1176LN, while other companies make their own FET compressors: Daking, Chandler, and Black Lion.

Add to cartLearn MoreAdd to list

Add to cartLearn MoreAdd to list

Add to cartLearn MoreAdd to list

VCA

Another compressor design showed up in 1976: the DBX 160, a compressor/limiter that used a Voltage Controlled Amplifier (VCA) housed in an integrated circuit for its gain reduction stage. Engineer David Blackmer had designed a transistor array with a -6 mV/dB logarithmic curve, meaning that every 6mV (millivolt) change in control voltage resulted in a signal level change of 1dB at the VCA’s output. Given that the other gain reduction approaches available (Variable Mu, optical, FET) were decidedly nonlinear, the dbx Gain Cell was noteworthy. Though Blackmer didn’t invent the voltage controlled amplifier — it had been used in synthesizers for some years — Blackmer’s VCA had a low enough noise specification to be suitable for pro audio productions. The dbx was a “hard knee” compressor; unlike Variable Mu and optical compressors, the onset of compression in the dbx 160 was instantaneous once the threshold was reached and was often audible. Though later VCA-based compressors can be quite transparent, some still love the distinctive sound of the older dbx units. dbx is even making a 500 series version, the dbx 650A. VCA-based compressors are often used on the mix bus, since the VCA’s control signal can be manipulated in exacting ways that allow precise settings of the attack and release times, the threshold, the compression ratio, and the knee to suit many styles of music. VCA-based compressors are made by API, Dangerous Music, and Solid State Logic.

Add to cartLearn MoreAdd to list

Add to cartLearn MoreAdd to list

Compressor Controls

Let’s look at some of the parameters used to control audio compression:

Threshold

The threshold is the point at which compression begins; when an audio signal is below the threshold, no compression takes place. Above the threshold, gain reduction begins. While many compressors have a threshold control, some others, including the iconic 1176, have a fixed threshold. The input control is used to determine how much signal is above the compression threshold. Here’s an example: If the signal going into the compressor peaks at 0dB and the threshold is fixed at -4dB, then the signal is 4dB over the threshold. If you raise the input level to +4dB, then the signal is 8dB over the threshold. Compressors that have a threshold control often don’t have an input gain control; with these models, if the input level peaks at 0dB, then you’ll lower the threshold control to -8dB in order to be 8dB above the threshold.

Attack

The attack time refers to the delay between the time when the audio signal crosses the threshold and the compression circuit begins to reduce the gain. Variable Mu and optical compressors usually have relatively slow attack times: 25ms for Manley’s Vari Mu, 10ms for UA’s LA-2A, compared to the 1176’s 20-microsecond attack time. While fast attack times will tame transients, if you lose the transients, then you can also take away the excitement of a track. It can be helpful to experiment with attack times (if the compressor offers the option) by starting with a slower attack — even as high as 50ms to 60ms. Then shorten the attack time to find the point at which the track sounds the best.

Release

It seems that release should simply be the reverse of the attack time — that is, once the signal drops below the compression threshold, the time it takes the compressor to quit compressing is the release time. But it isn’t always so; some optical compressors have a memory. According to the manual, an LA-2A has a release time of “60ms for 50% release, and then a gradual release over a period of 1 to 15 seconds to the point of complete release.” The time it takes for the LA-2A to fully release is dependent on both the amount of gain reduction happening before the audio falls below the threshold and the length of time that the audio had been above the threshold. But for compressors other than an LA-2A, it’s fine to think of fast attack and release times as being more “aggressive” and slow attack and release times as being “smoother.”

Ratio

The compression ratio specifies the amount of attenuation applied to a signal once the threshold is passed. The compression ratio is expressed in decibels, so a compression ratio of 4:1 means that a signal exceeding the threshold by 4dB will mean the resulting output will only increase by 1dB, so it’s attenuated 3dB — down to 1dB above the threshold. A signal that exceeds the threshold by 8dB will be reduced by 6dB, attenuated down to 2dB above the threshold. A milder ratio of 2:1 means that a signal that is 8dB above the threshold will be attenuated to 4dB above the threshold. In general, ratios of less than 4:1 are generally considered to be mild compression, from 4:1 to about 6:1 would be considered medium compression, 8:1 would be considered strong compression, and anything over 10:1 is generally considered to be limiting.

Knee

As discussed earlier, the original dbx compressors had a hard knee, meaning that as soon as the signal exceeds the compression threshold, gain reduction occurs at the selected ratio. Most others are soft knee, where the onset of compression actually precedes the threshold and starts gradually. Some contemporary compressors offer a choice of hard knee or soft knee.

Gain Makeup

Since the basic function of a compressor is to control the signal level by attenuating the loudest parts of the signal, most compressors have a gain makeup control to bring the overall level up after compression. The design of the amplifier section has a great deal to do with the sound of the compressor, whether it’s the “warmth” of the LA-2A or the “character” of an 1176.

Using Your Compressor

By becoming familiar with the sonic characteristics associated with the four main compressor designs — Variable Mu, optical, FET, and VCA — and the compressors which used those designs, along with the variables that those designs offer the end user, you’ll be able to develop a deeper understanding of the way that compressors work and how you can use them to your best advantage in the studio. 

Though few of us have access to racks of vintage analog compressors in the age of workstations and home studios, the information above is still useful because those vintage compressors have been modeled in the plug-ins you have in your workstation — and the software emulations are amazingly close to their hardware counterparts. Universal Audio still makes a hardware LA-2A, and its UAD version of the LA-2A behaves in the same manner — right down to some of the interesting nonlinearities of the optical circuit. The software versions of Fairchild’s 660 and 670 compressors are lovely, and if you want the aggressive response of an 1176, then you can find it in your workstation. 

I regularly choose the compressors I use in a mix based on my experience with the hardware versions of those compressors. For example, I’m familiar with the sound of a Variable Mu compressor, so if I want that sound and don’t have a Manley Vari-Mu available, then I can simply choose a plug-in based on a Variable Mu design. It’s the same with optical, FET, or VCA compressors; the emulations are quite good and are getting better with every revision. By learning the strengths and the weaknesses of each of the designs, you’ll be able to make better decisions when choosing the right compressor for any necessary task in the studio, whether it’s on a bass, a vocal, or on the whole mix. 

Sweetwater Has the Right Compressor for You

Are you still confused about which compressor you need? For help finding the right compressor, regardless of your budget or application — live or studio — you can call your Sweetwater Sales Engineer at (800) 222-4700, and they’ll help you decide from all the options.
The post How Do Compressors Work? appeared first on inSync.