One of the most misunderstood tools audio engineers use for live and recorded audio is compression. Over the years, manufacturers have tried to simplify it, but amateur engineers still try to overuse it — often with disastrous results. Here’s a primer on audio compression and how to use it effectively.
An audio compressor is designed to automatically reduce the level of peaks in an audio signal so that the overall dynamic range — the difference between the loudest sections and the softest ones — is reduced.
This “compressed” audio makes it easier to hear the nuance of the recorded sound, whether music or the human voice. Sometimes compression is called peak or gain reduction. This is because a compressor (or “limiter,” when acting more severely) rides gain on a signal much like a recording engineer does by hand when manually raising or lowering the faders of a mixing console.
A compressor’s circuitry automatically adjusts levels in response to changes in the input signal. It keeps the volume up during softer sections and brings the volume down when the signal gets louder. The amount of gain reduction is typically given in dB and is defined as the amount by which the signal level is reduced by the compressor.
Compression or limiting enables even the quietest sections to be made significantly louder while the overall peak level of the material is increased only minimally. The dynamic range of human hearing is about 120 dB. That’s the difference between the very softest passages humans can hear and the very loudest ones they can tolerate without severe pain.
Early recording media such as analog tape and vinyl offered much less dynamic range than today’s digital technology. This meant that using compression in the analog era was a necessity, raising the overall level of the material (making it “hotter”) without peak levels causing distortion.
While much of today's digital recording media approach or even exceed 120 dB of available dynamic range, quiet passages of recorded music can still be lost in the ambient noise floor of the listening area, which, in an average home, is 35 to 45 dB.
Compression is still important to digital recording. It can ensure that the digital signal is encoded at the highest possible level, where more bits are being used so that better signal definition is achieved. It also helps prevent clipping, a particularly harsh type of distortion that occurs when digital is overloaded.
In the real world of audio today, a vocalist may practically whisper part of a sound and then move to screaming within the same tune. Compression helps to level out the volume. It’s ditto for kick drums and bass, which can also change levels quickly.
Compression can also help to tame acoustic imbalances within an instrument. For example, these imbalances occur when certain notes of a bass guitar resonate more loudly than others, or when a trumpet plays louder in some registers than in others. Properly applied compression will make a musical performance sound more consistent throughout.
Compression is also used by broadcasters and voice-over narrators, whether working in professional announce booths or home studios.
Compression is needed today in virtually all recorded and live sound. Some modern mixers have a single control knob for compression in an attempt to make it easy for anyone to apply compression to the sound. But most professional users use a range of compression controls.
More sophisticated controls are threshold, which allows users to tell the compressor the level at which it can start reducing the amplitude of a signal. Ratio is a control that allows users to set the gain reduction with a ratio. For example, a 3:1 compression ratio means that if the signal is 3dB over the threshold set earlier, the output will be 1dB over the threshold.
Attack and release lets the user control how quickly the gain reduction starts and stops working. Knee allows the choice of how the compressor responds to signals that cross the threshold. Finally, output or makeup gain takes the compressed signal and boosts it to the right amount of volume to fit into the overall audio mix.
In order to operate, a compressor must first have some method of determining the level of the incoming signal and then control the gain using the fluctuations in that signal. There are many different circuit designs which have been developed to accomplish these tasks.
One of the most revered compressor designs is the LA-2A, a vintage tube device that was the brainchild of James F. Lawrence Jr., who had been a radar operator in World War II. Following his tour of duty, Lawrence began studying electrical engineering at the University of Southern California, while also quietly designing sub-miniature telemetry devices and optical sensors for the military.
But his passion was always radio, and he eventually landed a job as a broadcast engineer at KMGM in Los Angeles, where he soon became frustrated with having to constantly ride gain to ensure a proper signal. This led to Lawrence’s conception of a device he called a “leveling amplifier.”
Shortly afterwards, Lawrence started a company called Teletronix, setting up shop in his hometown of Pasadena, California in 1958. Among the line of broadcast products manufactured by Teletronix was a leveling amplifier — the LA-1 — of which about one hundred units were made.
Today, an advanced version of that design lives on in Universal Audio LA-610 ($1599.00), which combines a classic tube microphone pre-amp with the guts of Lawrence’s electro-optical element called a T4.
The T4 is comprised of a small light-proof metal canister that contains two components: an electro-luminescent (EL) panel (a device that lights up when electrical signal is applied) and a cadmium-sulfide photoelectric cell (a light sensitive device whose electrical resistance changes depending upon the intensity of light to which it is subjected).
It is the unique gain reduction characteristics that result from the interaction between these two components that predominantly gives an electro-optical compressor its signature sound.
The genius of the “el-op” circuit was its simplicity. The larger the signal that is applied to the EL panel, the brighter the light that is generated; the brighter the light, the less resistance the photo-cell (which controls the gain of the electrical circuit) exhibits.
Thus, the louder the incoming signal, the brighter the light and the more gain reduction is applied. This is done with virtually no harmonic distortion or audible artifacts. Conversely, when there is a small input signal (resulting in a dim light), the photo-cell will have a great deal of resistance and will therefore not affect the circuit at all, so there will be no gain reduction.
Perhaps the most important thing about electro-optical compression and limiting is that it is 100 percent program-dependent. That means both the degree of gain reduction and the compression ratio vary continuously with the incoming signal, making for a very natural sound.
Ironically, the first LA-1 became popular after first being used by Gene Autry, the singing cowboy. Autry’s use of the technology mandated its continued development. The technology went through several hands and was finally acquired by Bill Putnam’s company, later named UREI. Putnam’s son now runs Universal Audio.
The senior Putnam’s equipment was favored by artists such as Frank Sinatra, Ella Fitzgerald, Duke Ellington, Nat King Cole and Ray Charles. Much of his original gear remains in use today.
Though compression has a long and distinguished history in professional sound, it lives on today in both hardware (much of it vintage) and newer software plug-ins. Though compressors have become easier to use, the biggest problem remaining is the inexperienced engineer over using it.
Engineers should take the time to learn the basics of compression and remember one rule: use it conservatively. When a compressor is properly set, you won’t hear it working. Less is always more. Users need to accept that.
You might also like...
What’s old is new again. An ironic confluence of interrelated events — one that brought ribbon microphones to the front and center of broadcasting in the 1930s and to seemingly lose favor in the 1960s — is back again after 85 years.
Side-Bar to: Ribbons Microphones Make a Full Circle in Audio History (filed separately).
Unity was in the air at the 2017 NAB Show for real time production over IP networks, with broadcasters, vendors and standards bodies coming together for the IP Showcase.
Recording multi-track immersive audio is no longer difficult. Yet, many broadcasters and video producers prefer to stick with plain old mono sound. They are missing the opportunity to add major impact to their productions.
Since the world’s first audio recording in 1860, there have been legendary technical disputes in the field that are never settled. One more recent one is the question of which is better: digital plugins or hardware components? Debate is fiery…