Understanding Audio Dynamics: Part 1
In the world of music and sound production, audio dynamics and the control thereof is a core concept. And like most of the topics in this arena, the fundamental principles are actually very simple but they hide behind a vocabulary barrier. In this article, we break through that barrier to make audio dynamics and the equipment used to control those dynamics accessible and easy to understand.
The website http://www.dictionary.com offers two definitions for the word dynamic: “Change” and “Volume”. In the context of sound and music production, dynamic refers to “a change in volume over time”.
Dynamic range is simply the difference between the loudest part of a song and the quietest. One of the reasons that live performances provide an experience that can never fully be captured is the inherent dynamic range. Rock concerts and oboe recitals both traverse a range of loud and soft tones that simply won’t fit onto recorded media. The louds are too loud and the softs too soft. (Actually, the softest is silence and that CAN be faithfully recorded, but the difference between silence and the loudest parts, aka “the dynamic range” is what’s hard to capture). So, to we need to reduce or “compress” this dynamic range in order to record it.
If I were to be stranded on a desert island and only allowed to bring one dynamics processor, I’d bring a compressor and then I’d immediately sell it for bottled water and a flare gun, because who needs a compressor on a desert island.
Compressors are much simpler creatures than the marketing surrounding them implies. For example, here’s a quote from a recent advertisement for a new compressor plug in:
“The FRAM-a-STAT is a versatile dynamics processor that lets you shape the punch and tone of mixes with absolute accuracy. Its dual channel design functions as two separate mono channels. Using auto-makeup gain, you can adjust Threshold or Ratio with both “feedback” and “feed forward” compression types, switchable 3-position thrust high-pass filter, variable link channels and filter types, soft, medium and hard knee slopes and 6 Release settings.”
OK… I’ve been doing this since 1984 and I don’t know what half of that is supposed to mean.
The problem is this makes a compressor in general (and this one in particular) intimidating because they sound like a nuclear reactor when in fact, they’re just an automatic volume control.
“QUIET! You’ll wake the baby.” Compressors, Limiters, Expanders and Noise Gates in a nutshell.
Have you ever been watching a late-night movie and rushed to turn down the volume during a chase scene only to turn it up again to hear the whispered dialog?
Congratulations! You are a compressor.
Have you ever hit the mute button when a commercial started?
Congratulations! You are a noise gate.
And have you ever reached over in the car and cranked the volume knob further up when “your” jam came on? How about when you turn the radio down as the donation requests or commercials begin?
Congratulations! You are an expander.
These little illustrations sum up the major dynamics processors. The concepts are just that simple. Where the confusion starts is the vocabulary. Look at figures 1 and 2. Believe it or not, you’re looking at the same set of controls. One set we understand intuitively, the other seems confusing.
Lost in Translation
The “trick” here is to correlate the compressor’s controls to how you use your TV remote. The various dynamics plug-ins or hardware units all feature at least four key parameters which we’ll call “the big four”:
Let’s go back to the chase scene in that late-night move and use it to sort out these terms, starting with the compressor. And keep in mind that all a compressor does is to make the loud stuff softer and make the soft stuff louder.
As the chase sequence starts and the volume grows louder and louder, you start to become more and more uncomfortable. You and your spouse start exchanging worried glances (worried about waking up the baby or annoying the neighbors).
At some point you’ve had enough! It’s now officially too loud and you lunge for the remote control!
This “moment” correlates to: “Threshold”. Threshold is: the signal level at which the compressor becomes active.
How long it takes you to reach the remote correlates to: “Attack”. Attack is: how long it takes (after the threshold is reached) to begin reducing the signal level.
You’ve reached the remote and mashed down the “VOL DOWN” button and the movie is finally getting quieter.
How much you turn the volume down correlates to: “Ratio”. Ratio is the reduction factor by which the signal level is reduced. A mild ratio like 3:1 means a 3db increase of incoming signal will be output only 1db louder. In other words, a three-to-one reduction. High ratios like 10:1 or 12:1 are common in some dance music effects because they are noticeable and tend to create a pumping feel. More about that in part two of this series.
And if you set an extremely high threshold and set an “infinity-to-one” ratio, you’ve turned your compressor into a “limiter” which puts up a brick-wall and lets NO signal pass above that “limit” , hence the name.
Brick-wall limiters are commonly used as protection devices at the end of a signal chain to prevent blowing out speakers or overloading components. Alternatively, you can place a brick-wall limiter across your master output buss and use it to squeeze every last drop of dynamic range and musicality out of a mix.
The chase scene ends, presumably in a crash, and as the last hubcap is spinning down, the hero starts to whisper something melodramatic. You decide to start turning the volume back up, so you can hear what they’re saying.
This action correlates to “Release”. Release is: how long the compressor waits to return the signal to its normal level. Many units also provide a control for how quickly the level returns to normal.
All in the Family.
The other members of the Dynamics Family function in similar ways and by making use of similar (and sometimes identical) controls.
The noise gate works like a compressor standing on its head. Gates use similar controls but use them in a more extreme fashion (see figure 3).
With a noise gate, you set the threshold to determine when to cut off the signal completely. But unlike a compressor, which is looking for a signal that’s getting louder, a noise gate is looking for a signal getting quieter. And at a certain point (aka: the threshold) the gate closes entirely and cuts the signal off.
Noise gates, like their name implies, are used for removing noise and cleaning up a track. The quintessential application for a noise gate in recording music revolves around the drum kit. In a typical recording session, the drummer uses more microphones than the rest of the band (see figure 4).
A typical drum set up can have up to 12 mics. At a minimum you’ll have a mic on the kick, snare and hat. Usually you’ll have a pair of overhead mics to boot. And often you’ll have two snare mics, two kick mics and anywhere from two to four mics on toms. That’s a LOT of microphones for “one” instrument.
To clarify the use of noise gates, let’s take a close look at the microphones on the tom toms. They’re only really needed now and again during fills when the drummer actually hits a tom tom. But the rest of the time they sit there and soak up spill. The same goes for all the other mics.
And when you take that spill times X-number of mics? At best you get a drum mix that sounds noisy and unfocused. At worst you can introduce a ton of phase-cancellation issues.
Tip: If you ever have a case where the drum mix sounds “off”, but as you solo each track, each one individually sounds fine? It’s more than likely a phase issue. Try inverting the phase on some key channels like one overhead or the kick and snare and see if that tightens it up.
But an even easier solution is to avoid phase problems in the first place by turning off all those mics when not in use. And that’s exactly what a noise gate does.
Placed in-line with each of your drum mics, you can adjust the threshold so that the gate only opens (instantly and momentarily) when the tom tom is played. Afterward the gate immediately shuts off again and keeps all the rattle and spill out of your track.
Similarly, an expander is the mirror image of a compressor.
Expanders make soft stuff softer and can make loud stuff louder. This is a headscratcher for many people as the need for expanders is at first unclear.
I like to think of expanders (see figure 5) as a “soft gate”. They’re particularly well suited to taming headphone spill on vocal tracks. Like a noise gate, an expander waits until the target track has fallen below the specified threshold.
At this point the expander accelerates that fall by rapidly, but not immediately, necking down the volume until its silent.This works well on vocals because the slower reaction time tends to “ride” with the singer’s natural voicing where as a gate would be abrupt, would sound choppy and would interfere with the performance.
An Act of Creation
All of this being said, you can also press several dynamics processors into creative use as well.
If you listen to Madonna’s “Tell Me”, the rhythmic pules of the opening, and in fact the acoustic guitar throughout, is thanks to a very creative application of noise gates. Same goes for the voice, particularly in the opening few phrases.
Similarly, Katy Perry’s “Wide Awake” owes its strikingly “crisp” vocal bite to an exceptionally strong compressor set with about a 15ms attack. This lets the leading edge of each vocal phrase through at a much higher volume than the rest.
Listen to that song with your eyes closed and focus on the way “hard” sounds like the beginning of the word “Out” (in the line “Out of the lion’s den”) seem to crash through and grab you. Similarly, the “I” in “I was in the dark” does the same thing.
This is all thanks to some subtle but creative use of extreme compression. We’ll unpack exactly how to do this, and how to apply all the controls mentioned above, in the next article in this series.