This section is all about the tools we can use to change how something sounds. The trick is never overdoing it. It’s important to always remember with any of these effects; be it EQ, reverb, or anything else; it should be missing when it isn’t there, but not noticed when it is.
Equalization and Filters
EQ is one of the easiest to understand ways to change the way something sounds. Everyone has played with the bass and treble controls on their car or home stereo right? Even the novice can figure out how it works. With EQ, like everything else, less is more. Most problems that can be solved with EQ would be better solved by changing what mic is being used or where that microphone is. Additionally, it is usually a better idea to cut with EQ and not to boost; it’s easier to remove something undesirable than to try and create something desirable.
When it comes to EQ filter settings, there are 3 basic settings: gain, frequency, and width. Gain is the amount of boost or cut being applied; frequency is the point at which the boost or cut happens; width is the range of frequencies around the center that gets boosted or cut. (Width is often referred to in “octaves” or “Q”. 1 octave is a doubling of frequency and corresponds to an octave in musical notes [e.g. A4 is 440 Hz, therefore A3, an octave below, is 220 Hz.] A 2 octave filter is, understandably, 2 octaves wide; a ⅓ octave filter is ⅓ of an octave wide. “Q” is sort of the inverse of octaves; a higher “Q’ is a narrower filter and a lower “Q” is a wider filter. [e.g. a 2 octave filter has a “Q” of about ⅔, while a ⅓ octave filter has a “Q” of about 4.3. It’s a lot easier to think in terms of octaves and not in terms of “Q”, but many EQs use “Q” for filter width.] Width is a lot more complicated than it seems, but a full discussion how exactly width effects an EQ filter is far beyond what this reference seeks to achieve.)
A shelf filter applies gain to all frequencies above (for a high shelf) or below (for a low shelf) the set frequency. Often, shelving EQs don’t have a width setting, instead using a fixed width, but if they do, the width controls how “quickly” the gain rises before reaching the center frequency. Narrow width shelving EQs sometimes will “overcorrect” and give a boost at their center frequency, which can be used for effect.
Bell filters apply gain to the frequencies around their center according to their width. For example, a bell filter at 1kHz with a 2 octave width will roughly boost or cut from 500 Hz to 2kHz. Generally, to change the way something sounds (e.g. removing muddiness from a guitar), a wider filter is better, but to fix a problem (e.g. a microphone feeding back), a narrower filter is better.
High-, Low-, and Band-pass
These filters do exactly what they describe, although it seems a bit counter-intuitive at first. A high-pass (sometimes called a low-cut) filter will “pass” all frequencies higher than the center and cut all frequencies lower than the center. A low-pass (sometimes called a high-cut) filter will do the opposite, passing frequencies lower than the center and cutting the ones higher than the center. A bandpass filter combines low- and high-pass filters to only allow frequencies between the two center frequencies to “pass”. (A bell EQ filter is a variation on a bandpass filter that also has a variable gain.)
Several of these filters together is called a crossover, which are often used behind-the-scenes in speakers and system processors to make sure only the frequencies a speaker is designed to reproduce are sent to that speaker. In a two-way speaker, this ensures the smaller high-frequency driver doesn’t get overwhelmed by low-frequencies and the larger low-frequency drive only reproduces the low-frequencies it was built to handle.
The high-pass filter is mostly likely the one of all these filters to be used. If I have a high-pass filter with a variable frequency it is in use on every single channel to reduce the amount of low frequency noise in the system, leading to a much clearer mix.
Graphic EQs are a collection of bell filters, usually ⅓ octave widths, arranged to cover approximately the entire range of auditory spectrum, from 20Hz to 20kHz. These offer a rough visual representation of the EQ curve produced by them, though, with the advent of affordable digital mixers that clearly show the curve produced by a parametric EQ, graphic EQs are somewhat losing their usefulness.
This category is based on “reflections” — adding extra copies of a signal to change how it sounds. These methods imitate the same natural processes that make cathedrals sound like cathedrals and certain canyons yell back to you when you say, “Hello!”
Reverbs use a number of very short echoes — often too short to hear distinctly — to create the perception of a certain kind of physical space. There are a number of different kinds of reverbs, often called “Hall”, “Room”, or “Plate”. These are electronic simulations of what used to be real items. For instance, a “plate” reverb was originally a metal plate with a “speaker” on one end and “microphones” on the other. (To help you in your trivia knowledge, there are still actual rooms buried underneath Los Angeles studios with speakers and microphones in them.)
Most reverbs have an abundance of settings, from “length” to “pre-delay” and many more. These can be intimidating at first, but most FX processors come with presets to give you a good starting point. The more you use them, the more comfortable you’ll be with the various adjustable settings of each reverb.
Reverb is a great way to add depth to a mix. The more reverb used, the farther in the background something will sound. A good practice with stereo reverb is to pan opposite and a little wider to the “dry” signal. For example, if an acoustic guitar is panned 50% left, its reverb should be panned 75% right. This allows you to create space in your mix without sacrificing clarity.
Delay is very similar to reverb, except that instead of echoes indistinguishable from each other, delay uses discrete echoes to fill out a sound. I will often use delay instead of reverb with a very “full” voice, as delay will produce a similar result to reverb while not sacrificing clarity.
The most important settings in a delay are feedback and time. The time setting determines when the echoes will repeat. Delay usually sounds best when it fits musically, and most delays have a “tap” feature, which you can use to time your delay to the music. Used well, this means a delay can become a fill element of a mix, “repeating” the lead after a line and filling space in the mix. Most of the time, when I use delay I set the time to match the tempo of the music, unless I am using a stereo delay, in which case I use a “3-to-2” ratio between the left and right times, letting the delay “bounce” between the left and right speakers.
“Feedback” sets how much of the original signal is repeated each echo. When feedback is set to 100%, the delay is turned into a looper — continually repeating the original signal until it is lost in inevitable signal corruption. A feedback setting of 0% will repeat once. I usually like to set feedback between 10% and 50%; this gives you a repeating effect without overwhelming the original signal.
Where echo effects change a signal in time, modulation effects change a signal in frequency. This is achieved by either slightly modifying the pitch of the incoming signal, moving the signal slightly in time, or both.
A chorus effect simulates the sound of a choir. In a choir, many voices sing together, all slightly different, to produce a particular tone. One of the first artificial choruses was the céleste stop on a pipe organ, a series of pipes slightly out of tune from the others. Other examples are the piano, 12-string guitar, or mandolin; all of which use more than one string to produce a single note.
Chorus can be used to fill out a single instrument to make it sound “bigger” or to subtly cover up an instrument that is slightly out of tune by exaggerating its dissonance. It’s very effective on guitars and background vocals and, used sparingly, will sound quite nice on lead elements as well.
Phasers sound similar to choruses, but do so in a very different way. Instead of shifting pitch, phasers pass a signal through something called “non-linear” filters (like a small delay but not consistent across the frequency spectrum), intentionally creating phase problems to produce the desired effect. This interference behaves like a series of very incredibly narrow EQ filters, giving a “sweeping” sound.
A flanger is like a phaser, except the entire signal is uniformly delayed slightly to produce the desired effect. This gives a much more pronounced version of the phaser’s sound, turning the “sweep” into more of a “swoosh”. While they sound very similar, the measured frequency response of flanging is much more uniform than that of a phaser.
The third general kind of effect alters the signal in amplitude or volume. These can be used to “tighten up” a mix, but, used incorrectly, can cause unexpected problems as well. The two general types, compressors and expanders, work similarly to shelving EQ filters, only instead of working in the frequency domain, they work in amplitude.
These tools have the same basic settings, though sometimes the application is different. “Threshold” is the volume in decibels (dB) at which the compressor or expander starts to work. Ratio is the amount of compression or expansion done to the signal. Attack, Hold, and Release times define how long before the compressor or expander starts working, how long it continues to work when the threshold is no long reached, and how long it takes for the compressor or expander to stop working. “Knee” defines the severity of the compression or expansion. For example, a compressor with a “hard” knee will start working as soon as a signal crosses the threshold, whereas a “soft” knee will smoothly add compression as the signal crosses the threshold.
Compressor and Limiter
Compressors work by changing the volume of an incoming signal after it surpasses the threshold, according to the ratio. A 2:1 ratio at a threshold of -20dB means a -30dB input gives a -30dB output, but a -10dB input gives a -15dB output. The signal over the threshold has been “compressed”. Compressors often have a “make-up gain” setting as well, applying a boost to the output volume to “make-up” for the amount lost in compression.
A limiter is a special kind of compressor with an ∞:1 ration. This means that a signal will never get louder than the threshold level. Limiters are often used in loudspeaker protection, or as hearing protection in in-ear-monitors for just this reason.
Some compressors also include “negative” ratios. A 1:-2 ratio (sometimes labelled as -2:1) means the -10db input from the above example will output -15dB. While something to be aware of, negative ratios have very limited uses.
Compressors are a very handy tool for smoothing out a musician who doesn’t play consistently loud, or even tightening up the sound of a group of singers. A warning with compression, however: in live sound, a poorly set compressor is one of the leading causes of feedback. The make-up gain is usually the culprit here, as the extra volume from the makeup gain can easily push a system into feedback, usually right after the compressor stops compressing.
Gate and Expander
Expanders, and their more common sibling gates, work very similarly to compressors, but instead of acting on a signal above a threshold, they work on the signal below the threshold. For example, a 2:1 ratio at a threshold of -20dB means the -10dB input stays -10db at the output, but the -30dB input will become a -40dB output. Expanders often go into fractional ratios as well; continuing with the example, a “1:2” ratio means a -30dB input will become a -25dB output.
Like a limiter is a compressor with an ∞:1 ratio, a gate is an expander with an ∞:1 ratio. Gates will often have a “gain reduction” setting, allowing you to control how much, if any, of the input signal remains when the gate is “closed”.
Gates are one of the most useful tools we have to mix with. I use them to tame kick drums that are too boomy, reduce wind or handling noise in a microphone, hide a buzz coming from an overdriven guitar, and much, much more. If a gate is set incorrectly, then problems will arise. A gate with too high a threshold and too long an attack will leave the gate closed for the opening portion of a speech — and after any time the speaker pauses for longer than the hold time. The wrong gate settings on a snare mic will not open for a drummer’s ghost notes. While helpful, the wrong gate settings will quickly ruin an otherwise good mix.
There are many variations on compressors and expanders, such a duckers, de-essers, sidechain gates and compressors, multi-band compressors, serial or parallel compression, and more, but all of these are more involved than would be relevant here.
Saturation and Distortion
This last kind effect is somewhat strange to talk about, but it bears mentioning. Most of the time we consider distortion “bad”. We want the sound coming in to a microphone to sound like itself when it comes back out of the speaker. Distortion usually gets in the way of that. But just like a guitar overdrive can sound fantastic when used correctly, the same is true for other kinds of distortion as well.
When it comes down to it, every single thing we do adds a little bit of distortion to the signal we get. Microphones have their own distortion, preamps and mixers do as well, as do amplifiers and speakers. The trick is finding the “right” kind of distortion for the application. We often make decisions like these without really thinking about it: if one microphone sounds “better” than another, it usually means we prefer one kind of distortion to another.
I’m talking about very minor amounts of distortion here of course, but sometimes a more aggressive kind of distortion is valuable. One of the late Stone Temple Pilots vocalist Scott Weiland’s signature moves was singing through a megaphone; it created the sound he was going for.