Go into any music store and you'll probably find a glass display case containing a variety of little boxes with knobs, arrayed like so many candy bars. The boxes bear a wide spectrum of fanciful names—Distortion Plus, Mutron III, Fuzztain, Big Muff Pi, and so on. Some are no larger than a pack of cigarettes, while others could have anchored the Queen Mary. Some have the weight and feel of an empty tin can, while others seem ready to tangle with a bulldozer. The graphics on some are blatantly suggestive, while others sit in chrome-plated dignity, proclaiming their propriety to all who gaze upon them. These boxes are sound-modifier accessories. What are they, then? Of what use are they to synthesizer players? And how do you tell the good ones from the not-so-good?
Sound modifiers are most often built with guitar and bass players in mind. They usually have an input (axe) jack and an output (amp) jack and perform a single type of operation on the audio material being fed through them. More often than not they are battery-powered, and the battery is turned on when you plug something into them, a fact you should remember if you use one of them in a permanent setup. They can be divided into two general groups: those that are mathematically linear and those that are nonlinear.
Over-simply stated, linear devices do the same thing to a large-amplitude signal that they do to a smaller signal. Filters, phasers, and flangers are examples of linear devices. Linear devices are not supposed to introduce distortion or add frequency components that are not in the original signal. Of course, no signal processor is 100% linear, but we'll forget about that for now. Nonlinear devices generally change their mode of operation in response to the changing amplitude of the input signal. Distortion boxes, compressors, and ring modulators are all nonlinear devices. Devices in this class often exist only to create distortion.
Phasers are undoubtedly the most popular linear sound modifiers currently in use. These instruments work by actually shifting the phases (the positions in time) of the overtones, so that the higher-pitched overtones are shifted more than the lower ones. This in itself is not enough to produce an audible timbre variation. Then the phaser combines the phase-shifted signal with the original signal. In some frequency regions the two signals will cancel out, while in others they will reinforce one another. The result is that a phaser acts like a complex filter, emphasizing some portions of the spectrum and attenuating others.
The graph shows the frequency response of a typical phaser with the modulation shut off. The two peaks result when the phase-shifted portion of the output signal reinforces the straight-through portion. Notches between the peaks occur at frequencies where the phase-shifted signal is out of phase with the straight-through signal. Many phasers have emphasis or resonance controls that provide feedback from phaser output to input, thereby sharpening the response peaks. This is shown by the dotted line. (The conditions that give rise to phaser emphasis are the same as those that create resonant peaks in lowpass filter response. For a fuller explanation of the phenomenon, see my May/June '76 column.) The amount of phase shift is modulated at a sub-audio rate so that the emphasized bands sweep up and down through the spectrum. At low modulation speeds, phasers impart a gentle whooshing, jet plane effect to a static, harmonically rich sound. At higher speeds, a phaser-processed sound assumes a tremulant-like rotating speaker quality. Adding emphasis feedback changes the effect to a pronounced vocal-like quality.
The filtering action of a phaser is more complex than that of the filter section or any standard synthesizer that I know of. Not only does a phaser typically produce a multiplicity of spectral resonances and notches, but it imparts a motion to this complex frequency response. In combination with a synthesizer filter, a phaser is capable of fattening up harmonically rich sounds, especially those of the string family. If you happen to have the luxury of a stereo playback system, you can feed the phaser-processed signal through one channel and the unprocessed signal through the other. The effect will be more subtle but more spatial than if only one channel were used.
The amount of phase shift introduced within a phaser, and therefore the complexity of the resultant frequency response, is determined by the number of phase-shifting sections in the total circuit. Each phase-shifting section is capable of a maximum phase shift of 180°, which is one-half cycle of a sine wave. Four sections shift the phase from none (at very low frequencies) to two whole cycles (at very high frequencies). This gives a frequency response with one broad resonance and two notches, the minimum amount of spectral complexity for a phaser effect. Most popular units have six phase-shifting sections, which, as I mentioned before, gives two broad resonances. Eight- and ten-section phasers are available that give three and four resonant peaks respectively. Their effect, like that of a twelve-cylinder engine, is so smooth that it's hard to hear under some conditions.
In selecting a phaser, first feed the output of your synthesizer directly into a monitor amp. Set your instrument up so that it produces two triangular or sine waves (or filtered square waves) a semitone or so apart, two or three octaves above middle C. Listen for the low beat note, which is a result of intermodulation distortion in the synthesizer and/or the monitor system. Try to remember how much beat note you hear. Next, connect the phaser between the synthesizer output and the monitor amp input. Turn off the phaser modulation if possible. Listen again. You'll probably hear an increase in the loudness of the beat note. This is additional intermodulation introduced by the phaser. Now lower the output level of your synthesizer by turning its master volume control down. Find the point at which the phaser does not perceptibly add to the intermodulation-induced beat note. At this point, with the monitor gain set so that the synthesizer sounds right, does the phaser introduce an excessive amount of hiss or hum? If it does, then that particular phaser does not have adequate dynamic range.
This is a tough test for any signal modifier designed primarily for guitars. Guitar players generally like a little distortion, and accessory manufacturers often design in some "dirt" that seems to help guitar signals come to life. But it doesn't help keyboards! So once you've found one or more phasers that pass your distortion-noise test, then listen for the quality of the phasing to find the one you like best. There is no one thing to listen for, and there is no one "best" phaser. Sound quality in a device as complex as a phaser is a matter of personal preference.
In next month's column we'll talk about flangers.
For more articles by Bob Moog, please visit http://www.keyboardmag.com/analog/1318/moog-monday-hub-page/57761