Chorus, phasing and flanging are very common processes, but they don't always achieve the effect you hoped for. Paul White offers a few tips.
As followers of this series have probably gathered, rather than just explaining what effects do and where they might be usefully deployed, I'm looking at commonly used effects to see where they might cause problems. For example, we all think of chorus as a fairly harmless little effect that wouldn't say a cross word to anybody, but that doesn't mean to say that you can use it anywhere and be guaranteed a good result.
Chorus is created by adding a slightly delayed, pitch‑modulated version of a sound to the original sound, in roughly equal proportions. The intention is to create the illusion that two or more instruments are playing the same part at the same time. (Figure 1 shows block diagrams of all the effects discussed here.) The problem is that the illusion is only approximate, because if two or more people are playing together, any beat frequencies caused by tuning disparities will vary from note to note, whereas a chorus is modulated from an LFO, which produces a definite cyclic beating. Because of this giveaway cyclic pulsing, chorus has become an effect in its own right rather than an especially successful means of emulating ensemble playing, though if you pull an early string machine apart you'll probably find it contains three or more chorus sections, all running at different speeds to produce a richer, smoother ensemble sound.
Today, chorus is used extensively on guitars (thank Andy Summers for that, back in his Police days!) and occasionally on fretless bass, not to mention on keyboard pads. It creates a sense of movement and stereo width, but those very attributes also tend to push a sound further back in a mix. If you want a sound to stand out, it should be fairly dry (in terms of added reverb), and it should be positioned close to the front of the mix. Chorus, however, has the effect of 'de‑localising' a sound — it sounds rich and wide, but you don't really know where it's coming from, and the psycho‑acoustic outcome is that it sits further back in the mix. If you pull up the sound's level to try to get it to the front again, it may become too loud, so the most sensible way to use chorus is to apply it mainly to those sounds that are supposed to be part of the backdrop. This is only a general rule and it can be thrown out of the window when you've got a fairly sparse mix — which is why chorused fretless bass works well in slow, moody music and chorused guitar helps fill out the sound in a small band.
Variations On A Theme
If you want to add life to a keyboard sound, alternatives to chorus include pitch detuning and Leslie (or simulations), as well as layering sounds and then detuning one of them slightly. Where some chorus effect is desired to add interest to a sound, but you don't want to the sound to recede in the mix, try a very shallow, slow chorus where the cyclic effect isn't too obvious. I've also found that panning a mono chorus to one side of a mix and the dry sound to the other often produces a more spatial effect than a dedicated stereo chorus unit, and without making the sound seem so disembodied. Watch out for early stereo chorus units that achieve their effect by phase inverting one of the outputs. As soon as you play your mix back in mono, the effect will disappear completely! Test for this using the mono button on your desk, and if you find you have such a unit, use just one output panned hard to one side and the dry signal panned hard to the other, as just described.
Multitap chorus effects are less monotonous‑sounding than simple chorus effects (though they can sound even more disembodied), and a useful alternative to the traditional way of working is to add chorus only to a track's reverb send, so that the dry sound stays solid and up‑front, while the reverb takes on an attractive shimmer.
In situations where you specifically want an ensemble effect, try a pitch‑shifter set to detune mode rather than a chorus effect, as this will avoid the regular beating effect. As a bonus, most pitch‑shifters also include a variable delay time, to emulate the effect of two players not quite in perfect time with each other.
In my own experience, chorus doesn't work very well for thickening vocals, and it doesn't do much for percussion either. Vocals are better treated with a pitch detuner or a dedicated ADT (Automatic Double Tracking) effect, while drums can be put through with a flanger if a special treatment is needed.
Chorus creates a sense of movement and stereo width, but those very attributes also tend to push a sound further back in a mix.
Going Through A Phase
Phasing was the next purely electronic effect after the wah‑wah and fuzz box, and it originally set out to emulate tape phasing (history has since called this flanging). In the phaser, a signal is again split onto two paths, one straight and the other diverted via a phase‑shift network that can be varied under the control of a low‑frequency oscillator. The result is a continuous tonal change based on the frequency cancellation that occurs when the phase‑shifted signal path is out of phase with the straight signal path. Because the cancellation process is frequency dependent, the process creates a deep notch in the frequency response of the output, and under LFO control this notch moves across the audio spectrum, creating a tonal sweep. Subjectively, the sound is far less strong than flanging, as there's no significant delay between the two signal paths, but a number of guitar players latched onto the sound during the psychedelic era, and it became part of the guitar effect repertoire. Exponents include Jimi Hendrix (in his more laid‑back songs) and Pink Floyd's David Gilmour.
As with chorus, phasing has the effect of taking away the immediacy of a sound, making it sound disembodied — indeed, that's why it became so popular for psychedelic music. The ethereal nature of the sound is emphasised by the addition of reverb or chorus, but to stand out properly the effect needs to be used in a sparse musical arrangement that has plenty of space. In a busy mix, the phased sound will slither around in the background unless it's mixed very high.
Phasing works best with harmonically rich sounds, so overdriven guitar is ideal — just remember to put the phaser after the overdrive. The effect may also be used on some synthesized sounds, to create movement, and even though it does tend to push a sound to the back of a mix, the ever‑changing tone means that the listener is more likely to be aware of its presence than an instrument that has a constant sound.
Chorus and phasing might be polite little effects, but flanging can have real attitude. It came onto the scene a while after phasing, and employed the then brand‑new technology of the charge‑coupled analogue delay line to produce the amount of delay necessary to emulate tape flanging more closely. As with phasing, two signal paths are used, one dry and one processed, but this time the processed sound is delayed by several milliseconds, as well as being modulated. The underlying technology is almost identical to that of chorus, except that chorus tends to use slightly longer delay times and doesn't feed any of the output signal back to the input. A flanger creates its deep, almost resonant whooshing effect by feeding some of the delay output signal back to the input, and although this has no counterpart in manual tape flanging, the effect is pretty dramatic. If you were to look at the spectral characteristics of the output signal, you'd see a whole series of strong peaks and notches in the response; these move across the audio spectrum under the control of the modulating LFO.
Because the effect of flanging is so distinctive, it's unwise to over‑use it.
Flanging has no direct counterpart in nature — other than, possibly, the sound of a jet plane passing low above a city with lots of reflective concrete walls — and because the effect is so distinctive, it's unwise to over‑use it. Like phasing, flanging works best on harmonically rich sounds, but it is also strong enough to show up clearly on clean guitar, drums, or even vocals. The real problem is that flanging is a cliché, so finding artistic justification for its use isn't always easy. One tip that might help is to put a flanger before a reverb or delay, as this can provide interest without making a sound seem over‑effected.
On the technical front, many people also complain of unwanted distortion being caused by flangers, even though the input meters are reading OK. This happens because of the large amounts of internal feedback used, and because the input level is monitored before the feedback, there can be situations where the internal circuitry distorts at levels far below maximum on the input meter. The answer is to ignore the meter and keep reducing the signal level until the sound cleans up.
Another technical problem with flangers is noise — with all that feedback the flanger is perched on the edge of instability, so even when there's no signal you can still hear flanged background noise being produced. Where possible, the flanger output should be gated before being passed into delay or reverb effects, but in a multi‑effects unit with a fixed internal routing structure, this may not always be possible. In mitigation, though, the flanger noise is itself flanged, so it's more likely to be perceived as part of the desired effect, except during pauses. If you can arrange to mute the flanger during pauses, you can get away with even quite a noisy unit. As with chorus, check that older stereo units really are stereo and that they don't simply phase‑invert one of the outputs.
Next month, I'll be looking at some of the trouble you can get into with delay and echo effects.