Focus on the sound, not on the tool: Gate terminology and how to use it.
Jun 05 2015
We recently encountered someone who’s been using gating terminology exactly backwards. Yet, from their perspective they were using it “correctly,” and have been for a number of years. (Luckily, they are an user of audio processing, not a tech coach!) Once we sorted it out, it was interesting to see the logic behind their misunderstanding. It turned out to be a lesson that goes far beyond the mechanics of noise reduction. It reflects a principle of good production overall.
To get everyone up to speed, let’s define what an audio-processing “gate” is:
A gate is an audio-processing tool that eliminates noise between words. It does this by allowing louder sounds (such as your voice) to be heard, and softer sounds to be silenced. Using the gate’s one main adjustment, called a “threshold,” you set the level between your softest wanted sound (such as the very end of a word) and the loudest unwanted sound (such as room noise, computer fan, soft breaths, or low-level mouth clicks).
In the case of room tone, your first thought should be to eliminate or reduce any noise from a computer’s fan, ventilation hum, minor hiss, etc. But no room is 100% silent, except for hugely expensive test chambers. Ordinarily, that little bit of remaining background noise is hidden or masked by your voice, or music, etc. … or at least, the casual listener is distracted from hearing whatever small level of noise exists. But when you are not speaking, such as between sentences, the noise can become apparent, along with those mouth clicks, breaths, etc.
That’s where the gate comes in. If set correctly, it works very nicely.
But if set incorrectly, it can cut off the tail of your last word as you exhale, or can make hiss or room tone conspicuous by its sudden absence. Or, after the pause, as you begin to speak again, the gate might not cease it’s “quieting” effort quickly enough, so that the start of your first word is cut off, or begins very abruptly.
As the tool’s name suggests, it’s a gate that opens and closes. It “opens” to let the sound come through. And that’s where our friend’s misunderstanding comes in.
The point where the gate opens (enabling sound to be heard)is called its “attack.” The point where the gate begins to close is called the “release.”
To our friend, that didn’t seem logical. After all, the gate is attacking the sound, no? Wrestling it to the ground , as it were. And, when you begin speaking again, he thought of it as releasing its grip, as a captor would release a hostage.
In fact, that interpretation is perfectly logical, IF your focus is on the action of the gate. But that’s what’s wrong. Focusing on the tool is the wrong perspective in skills from acting to carpentry.
Instead of focusing on the gate and its function, our focus should be on the sound. Fix the perspective, and the correct interpretation makes equally good sense. The word “attack” has same meaning as it does to a musician – the attack on, say, a trumpet note involves how the note is initiated. Is it a soft, gradual attack? Or a hard, sudden one? As with the trumpet, the “attack” part of a gate’s operation is when the sound begins to be heard. That is, when the gate begins to open.
Similarly, “release” doesn’t refer to the gate’s hold on the sound, it refers to the sound that is being released — as when you release your breath at the end of a word. Or, again, the way a musician releases that trumpet note, or lifts the finger from a piano key.
(By the way, don’t take this analogy too far – the term “decay” is also used in describing a musical tone, but in audio recording and processing it can mean something different; the word “decay” isn’t part of gate terminology. In voice recording, you might think it means the decay of an echo or reverberation, as the echo falls off and disappears, and you’d be right. But when talking about an instrument’s sonic character, “decay” means the inherent lowering of, say, a guitar string’s volume right after it’s plucked. The string’s sound is then “sustained” until the string (and thus the sound) is “released.”
Which brings us back to our subject … )
Attack and release points are shown in this graph of a gate’s operation at Wikipedia: https://en.wikipedia.org/wiki/File:Noise_Gate_Attack_Hold_Release.svg. Notice how the audible portion of the sound – not the gate – is at center of the chart.
Also notice that the gate typically is not set to instantly open and close. The release and attack phases are more or less gradual, depending on the nature of the sound and the effect desired. In processing a voice over recording, it’s generally best to be conservative, in the sense that the gate shouldn’t act too abruptly in either direction, and sometimes it shouldn’t be fully shut. That is, you may want a little bit of background or breath sound to come through, rather than an eerie silence. Some gates allow you to set the amount of attenuation (aka the gate’s “range”). Depending on your software, there may also be other settings available in your gate.
How to determine what settings are best for your recording? That’s the point, and it applies to all aspects of recording and production. Use your ears. Beyond the very basic stuff like properly setting volume, focus on the sound, not the settings.