Home -> Magazines -> Issues -> Articles in this issue -> View
49 Hot Signal Processing Tips | |
Article from Sound On Sound, June 1989 |
The more imagination you apply to signal processors, the more you can get out of them. Craig Anderton presents 49 tips that take you beyond stock effects and into new ways of creating your own sonic signature.
Signal processors can make a dull sound shine, a dry sound wet, and an uninteresting sound fascinating. Processors can make a synthesizer sound more like an acoustic instrument, an acoustic instrument sound more like a synthesizer, or make either sound like nothing you've ever heard before.
The flimsy, unreliable foot pedals of yesteryear have grown into sleek, rack-mountable, MIDI-controlled, programmable 'do-all' boxes that cost less and perform more efficiently with every passing year. Not too long ago, £200 would buy you a horrible-sounding, analogue delay-based echo unit; now it buys you a digital reverb. Yes, we've come a long way.
Yet so much of the potential of signal processors remains untapped. It's a temptation to find a couple of favourite settings and rely on those (especially today, when programming a new effect from scratch can take some time). But deep within a signal processor lie new settings and applications waiting to be found.
The object of this article is to examine seven of the most popular areas of signal processing, from digital reverb to multiple effects systems, and come up with seven useful tips for each one. Some of these tips are pretty basic and some are relatively esoteric, but all of them are designed to help you get more out of your signal processing gear. So plug in a few patch cords, feed in some source material, and let's get started.
People tend to think of reverberation as a 'stock' effect that you just plug in and use, but reverb can benefit from additional signal processing as much as any other electro-musical device, so let's talk about some of the most popular ways to combine signal processing with reverb.
1. Reverb followed by equalisation (EQ).
Equalising the reverb's output can enhance the overall sound as well as more closely approximate the sound of an acoustic space.
For example, boosting lower frequencies gives a larger sound, since in a large acoustical space, high frequency echoes will fade away sooner than lower frequency echoes. Thus, the reverb sounds more distant. Boosting the higher frequencies gives a bright echo sound with more presence, which subjectively makes the reverb appear closer.
EQ can also help offset some limitations of particular reverbs. Inexpensive digital reverbs can lack exceptional high frequency response, so boosting the treble a bit after the reverb can give a little more 'sheen'. (Note that boosting treble prior to going into the reverb may reduce the overall dynamic range, which is the reason for patching EQ after the reverb.) If you want a less bassy sound, put the EQ after the reverb and reduce the low frequency response.
Another EQ application involves notching out specific frequency ranges of the reverb signal. For example, to make a vocal really stand out in a track, reduce the reverb's midrange response so that the reverb doesn't compete with the midrange-rich vocal.
2. Equalisation before reverb.
There are some reasons for patching EQ before the reverb. Since reverbs typically have a limited dynamic range, there's no point in putting any more signal into a reverb than necessary. So to leave more dynamic range for other frequencies, reduce the bass (or treble) going into the reverb.
3. Delay prior to reverb.
Although many digital reverbs include pre-delay (ie. a slight delay before the reverb begins) as part of their arsenal of reverb effects, this will often not be included with special effects such as gated reverb. Adding a relatively short (30 to 80 millisecond) delay before the signal hits the reverb will add a feeling of space to any reverberated sound, as a sound in a physical space takes some time to reach the room surfaces and bounce off them to create the reverb effect.
4. Limiting before reverberation.
The easiest way to make sure that signal levels don't exceed the reverb's dynamic range (which could create distortion) is to patch a limiter just before the reverb unit. Since a limiter can clamp signal levels to a maximum, presettable limit, it will prevent distortion from occurring no matter how much you increase the signal level going into the reverb. Careful, though: too much limiting may give a 'squeezed', unnatural sound.
5. Chorus after reverberation.
If used subtly, light chorusing added to reverberation can impart a shimmering, full sound that is quite pleasing. To achieve this, set the chorus for a delay in the 20 to 30ms range, with just a little bit of modulation. Shorter delays produce flanging effects, and while flanging can sound good, if not handled carefully it can also sound gimmicky.
6. Reverb 'splashes'.
Putting some kind of level control or footpedal in front of a reverb unit allows for reverb 'splashing'. This technique lets you selectively add reverb to specific notes, phrases, solos, etc. The basic idea is to keep the pedal or level control at a fairly low level (or off altogether). When you want to add reverb to a particular sound, turn up the level control just before the sound occurs, then turn the level control back down right after the sound has happened. Only the sound you want reverberated will be reverberated.
7. 'Chopped' reverb.
It never hurts to experiment with signal processors, and one of the more interesting patches I've tried in a multitrack environment is something I call 'chopped reverb'. Patch the reverb output into a noise gate and trigger the gate from a taped, snare drum track. This synchro-sonic type of effect gives a little shot of reverb every time the snare drum hits, which adds more rhythmic unity to the sound.
People who use flangers only for flanging are missing out on a number of useful applications. Here is a list of such applications and recommended control settings to 'put you in the ballpark' regardless of what kind of flanger you use.
1. Vocoder simulation ('robot voices').
While nowhere near as versatile as a real vocoder (in my opinion, one of the most underrated signal processors around), the following gives a pretty good simulation of vocoded vocal effects.
Select negative feedback (if possible), and turn it up just short of the point of regeneration. Set the mix control to delayed sound only, with no modulation. Now, slowly increase the initial delay time (or, if you have only a few preset settings, go for a delay time in the 15 to 20ms zone). Your voice should take on a mechanical, robot-like sound; if not, turn the feedback control up further.
2. Vibrato.
Set the blend (mix) control to output the processed (delayed) sound only; otherwise, you'll hear flanging. Then, crank the modulation depth up full (with the modulation speed set around 7 to 10Hz) and adjust the initial delay for shorter and shorter delay times until you get just the right vibrato depth.
Selecting a short delay time is important because many flangers are still based on inexpensive analogue delay technology. An analogue delay line's fidelity improves with shorter delay times, and since we're listening to the delayed sound only, it's important that we have the best possible sound quality. Also, the modulation depth will usually decrease with shorter delay times, so what might sound like wildly out of tune pitch shifting at longer delays becomes a pleasant vibrato effect at shorter delays. A final note: for best results, don't add any regeneration.
3. Fast rotating-speaker sound.
Flangers don't produce convincing slow rotating-speaker effects, but for fast rotating-speaker sounds, use the settings given above for vibrato but select an equal blend of delayed and straight sounds. You'll hear that familiar 'bubbly' type of sound.
4. Phase shifter simulation.
While flangers and phasers are quite different, it is possible to achieve a good imitation of a phase shifter with a flanger. Since phase shifters don't give much of a time delay and can't sweep over as wide a range as a flanger, the trick is to set the initial delay fairly high (say, 5ms or so) and use only a moderate amount of modulation depth.
5. Equaliser effects.
A flanger can also serve as a comb filter, which adds a series of frequency response peaks and dips to a signal. This requires minimum modulation depth to prevent any cyclic sweeping. You can control the effect in three ways: varying the initial delay control changes the sound's overall frequency response; varying the blend alters the depth of the filtering effect; and varying the regeneration changes the peakiness of the equalisation. While a flanger is no substitute for standard graphic or parametric equalisers, it can still give some novel effects that the more common equalisers cannot.
6. Stereo simulation.
This one is a little tricky to get right, and you'll have to split the signal into two channels: one into the flanger and one straight into an amp (see Figure 1). However, if you're playing through two amps or want to take a mono track from a tape recorder and make it stereo, this is a useful technique.
Set the flanger for no modulation, delayed sound only, and no regeneration. As you increase the initial delay control, you'll widen the stereo image until, at longer delay settings, you'll hear a slapback echo. The delayed sound may not be quite as hi-fi as the dry sound, so you might want to lower the level of the delayed channel just a bit. Although this shifts the stereo image towards the dry channel, the increase in fidelity may be worth it.
7. Through-zero flanging.
True tape flanging varies the time difference between two signals from zero milliseconds upwards, but electronic flangers have a finite minimum delay time - typically a millisecond or so, preventing them from creating the wide range flange effect of tape-based flanging.
For true through-zero flanging, you'll need two flangers (preferably of the same make and model) hooked up as shown in Figure 2. Set flanger 1 for minimum delay and delayed sound only. This now becomes your 'dry' signal. Set flanger 2 for delayed sound only, adjust the modulation for a suitable sweep, and set the feedback on both units for the same amount. When flanger 2 reaches its minimum delay, there will be a true zero time differential between the two flangers, thus producing through-zero flanging. Since the minimum delay is so small, you probably won't notice any time lag between the slightly delayed 'dry' signal and the real-time signal being processed.
The parametric equaliser is a tone-shaping device that typically includes from one to four filter stages, each offering variable frequency, bandwidth (also called resonance or 'Q'), and boost or cut at the selected frequency. These stages are sometimes connected in series to provide processing at different points in the audio spectrum for a single signal, but there are usually patch points so that each filter stage can serve as a single-channel equaliser for multiple signals.
Because parametrics alter frequency response in an extremely precise way, they are often used for solving specific problems instead of general tone-shaping (for which graphic equalisers are better suited). Here are seven ways to solve problems using a parametric.
1. Eliminating resonance problems when recording acoustic instruments.
Once, when recording classical guitar, I found an extreme frequency response peak at around 200Hz. Whenever the guitarist hit a note in that region, the recorder would distort, and setting the levels to keep this particular frequency below the distortion point made the rest of the guitar's frequencies too low. A compressor or limiter could have solved this problem, but only at the expense of reducing the sound's dynamics. A parametric EQ saved the day. I simply programmed in the resonant peak's frequency, cut the response at that point, and adjusted the bandwidth so that the amount of notching precisely offset the peak. The end result was a signal that recorded much better on tape, didn't have any 'boominess', and didn't sound compressed.
2. Eliminating hum.
Suppose you've recorded a guitar track that's perfect except for a bit of 50Hz mains hum. No problem: set the parametric's frequency to 50Hz, select the narrowest possible bandwidth, and go for maximum cut. The hum will disappear, yet the rest of the signal will be virtually unaffected.
3. Fixing 'dead spots' in instruments.
Some instruments, particularly bass guitar, will have some notes that don't respond as strongly as other notes. Adding a little bit of boost at these frequencies creates a more even response. This same problem can occur with amplifiers, where there might be a response anomaly in the speaker. Again, a little boosting (or in some cases, cutting) will solve the problem.
4. Salvaging premixed tape tracks.
In many 4-track tape setups, it is necessary to premix several existing tracks in order to make room for overdubs. However, what sounds like a good mix at the time might not seem quite so good when all the other parts are added. A parametric can do tricks like reduce just the kick drum, if it is mixed too high, or reduce some of the midrange of the premixed track if it competes with vocals overdubbed later.
5. Reducing feedback problems with PA systems.
Feedback in PA systems tends to occur at certain specific frequencies. By placing a deep notch at those frequencies, with one or more parametric stages, the PA will be less prone to feedback problems and you'll be able to get more level out of the system before feedback occurs.
6. Removing clock bleed-through from older, time-based effects.
Analogue and early digital delay lines often produced a faint, high frequency whine at longer delay settings. To remove this, simply set the parametric's frequency to the frequency of the whine, then select the highest possible resonance and maximum amount of cut.
7. Creating response peaks in synthesizers.
Most acoustic instruments tend to have response anomalies (called 'formants') that add character to the sound. Synthesizers, on the other hand, generally have a uniform frequency response throughout the instrument's range. Some musicians feel this results in a flatter, less interesting sound. If you use a parametric to add some artificial peaks and dips in the synthesizer's response, the resulting timbre will more closely resemble that of traditional instruments.
Multiple effects processors do so much more than one signal processor by itself, and some combinations of processors are particularly pleasing. Let's look at some of the most popular multiple effects combinations, designed mostly with guitarists in mind.
1. Compressor followed by fuzz.
This is one of the most popular effects combinations for guitarists who want more sustain. The compressor raises the volume of a string as it starts to decay, thus increasing the apparent sustain. This not only makes for a smoother-sounding fuzz, but most compressors can provide a volume boost should you want to overload the fuzz for even more distortion. A further improvement is to follow the fuzz with a noise gate or volume pedal, as this can mute the system to cut out noise when you're not playing.
2. Phaser followed by phaser.
One problem I've found with phase shifters (or flangers) is that the repetitive, cyclic nature of the phasing sweep can become pretty monotonous. So, here's a way to add more 'animation' to a phase-shifted signal. If you set each Phaser's LFO modulation asynchronously, the swooshing effect becomes more random and less repetitive. Start by setting one LFO speed fast and the other slow; this gives a slow, repetitive sweep that has a 'bubbly' or 'tremolo' sound superimposed on it. Another option is to set both LFOs at moderate sweep frequencies (slower than vibrato, but fast enough to be noticeable), then alter one sweep setting slightly so that each phaser has a slightly different sweep. This produces the most randomised effect.
3. Stereo delay.
Two delays (Figure 3) can definitely sound better than one, especially if they're set to different delay times. Set both units for the same proportion of delayed and dry sounds, as this will place the dry sound in the centre of the stereo image. I've found that setting the shorter delay to between 60% and 80% of the longer delay time works very well; for starters, try adjusting the longer delay to about 150 milliseconds, and the shorter delay to about 100ms.
4. Taming the envelope-controlled filter.
Envelope-controlled filters are underrated little devices, but are sometimes too extreme in their effect. For example, if you try most envelope-controlled filters with bass or drums (and by all means, try them on drums), the tone will be unsatisfactorily thin.
The solution (Figure 4) requires a Y cord and two-channel amp. Simply split the bass or drums, and use channel 1's level control to set the level of the unaltered signal while channel 2's level control dials in the desired amount of effect.
5. Fuzz before flanger/phaser.
Flangers and phasers give the most intense effect when they have a signal with lots of harmonics (overtones) to work with. Adding a fuzz makes for a harmonically rich sound that accents the flanging or phasing effect.
6. Pre-delay.
Although generally thought of in the context of reverberation, delaying a signal prior to chorusing, flanging, filtering, or envelope filtering can help diffuse the delayed signal a bit and put it more in the background, compared to the 'dry' signal. Figure 5 shows a setup using a two-channel amp, but you could use two separate amps for stereo, instead.
Match the delay time to the tempo of the music for an even more intriguing effect. By the way, make sure you try feeding a digital delay set for long, repeating echoes into a digital reverb; the effect can be wild.
7. Post-delay.
A variation on the above effect is post-delay, where you add delay after a signal processor. Figure 6 shows one of my favourite post-delay patches, great for vocals, needing only an equaliser and delay line. Set the graphic EQ for minimum lows and maximum highs, so that only the highs are echoed, giving a bright sound that doesn't muddy the original signal. It's better to add the delay line after the graphic, because the equalised signal will have less level, and you'll be able to pump more signal through the delay than if you feed the full-bandwidth signal into the delay and then EQ it.
Rack mounted guitar preamps typically include a preamp, output stage capable of interfacing with studio/PA mixers or standard guitar amps, fuzz, effects loop, and equalisation. A rackmounted guitar preamp, combined with a power amp and suitable speaker system, produces a setup that is more flexible (and more easily serviced) than a standard, all-in-one guitar amp.
The following control settings are intended as points of departure for particular musical styles. Remember, though, that individual guitars, guitar control settings, styles of playing, and amplifiers all affect the sound. For example, if you have a 'dirty' amplifier, then the fuzz effects added by the preamp might be less obvious. For this reason. I'd suggest using a clean power amp and conservatively rated speaker system with any rack-mounted preamp. This way, your sound is totally controlled at the preamp, not by other variables in the power amp and speaker system.
Although different manufacturers use different names for the same functions (why is that, anyway?), most rack-mount preamps will include the following functionally equivalent controls. The fuzz or overdrive section will typically include a sensitivity control to adjust the intensity of the overdrive effect, as well as a control for setting the overall fuzz level (when cutting in the overdrive effect, this control lets you select a fuzz sound which is louder, softer, or the same level as the non-fuzzed sound). The equalisation section will typically include, at the very least, bass, midrange, and treble controls; there might even be an additional midrange control, and switches for high or low frequency cut (reminiscent of the rumble and scratch filters found on hi-fi preamps) or for high frequency boost (usually called 'brightness'). There will also be some kind of overall volume control, which is adjusted to match the preamp to the next stage in the signal chain.
By the way, if you don't have a rack-mount preamp but do have a combination of effects that includes fuzz and equalisation, or an all-in-one amp with fairly sophisticated control capabilities, the following tips should help you obtain specific sounds.
1. Jazz sounds.
Most traditional jazz guitar players covet a clean, bassy, warm sound. So, start off by making sure that the fuzz section is out of the signal path. Then, start working on the EQ: turn up the bass to provide a strong bottom, keep the midrange flat (or slightly boosted if the midrange frequency point is relatively low-around 500Flz), and pull the treble back a bit to take any 'edge' off the overall sound.
2. Country EQ setting.
Country guitarists tend to go for a bright, twangy sound without distortion. So again, leave the fuzz section out of the signal path and concentrate on EQ. Generally, you'll want to trim the bass a little bit (but not too much), accentuate the upper midrange (around 3kHz) to taste, and boost the treble considerably. Because country music places an emphasis on vocals, guitarists generally don't play all that loud. A bright sound lets the guitar cut through, even at relatively low volumes.
3. Simulated acoustic guitar setting.
In some cases (say, for rhythm parts), you may want an electric guitar to sound as much as possible like an electrified acoustic. Keep distortion out of the picture again and turn the midrange way down, boost the bass a bit for a more full bottom end, and boost the treble for a bright sound that accentuates the plectrum. De-emphasising the midrange 'leaves room' for vocalists and other midrange instruments, while boosting the treble gives a bright sound that cuts through a track, even at low volume levels. I'd also recommend using the guitar's bass (rhythm) pickup by itself, although using both bass and treble pickups can work well.
4. Funk rhythm guitar setting.
In most funk music, the extreme emphasis on bass and drums implies that guitar parts must be bright and clean for balance. Using the guitar's bass pickup does this, but if you're looking for a really thin sound, try throwing the bass and treble pickups out of phase with respect to each other. Distortion doesn't work here either, so keep the fuzz out of the circuit. For EQ settings, leave the bass flat or slightly trimmed, trim back the midrange a bit (leaving room for the vocals), and turn up the treble. If you have a brightness switch, turn that on, too.
5. Heavy metal lead setting.
This is where the fuzz finally comes into play. Use the bass pickup for the smoothest lead sound, or the treble pickup if you want a brutal and bright lead effect. For maximum sustain, patch a compressor either between the guitar and preamp or, if there's a pre-EQ effects loop included in the preamp, patch the compressor into that. Adjust the fuzz as desired, but you'll probably want the sensitivity up full or close to it. And while you're at it, turn the bass control up for a rumbly bottom, leave the midrange flat (or add a sparing amount of upper midrange boost), and adjust the treble to obtain the degree of 'cut' you need. Generally, the fuzz itself will generate enough high frequencies that you'll want to pull back on the treble a bit to keep a balanced sound.
6. Blues leads.
Blues guitarists tend to go for a more sharp, defined sound than heavy metal types. Some fuzz is indicated, but not too much: you want a hint of distortion, not a wash of overload, and a little compression helps give sustain. I'd suggest turning the bass up a lot, the midrange up a little, and keeping the treble pulled back a bit (again, if you need more 'cut', boost the treble more).
7. Old amp setting.
This is designed to give that Keith Richards, amp-with-a-zillion-miles-on-it, dirty rhythm sound. I'd recommend using the bass pickup. The crucial factor is to set the fuzz so that distortion occurs only on the peaks of your playing; this simulates the sound of the amp breaking up. As far as equalisation is concerned, pull back on the bass and treble and go heavy on the midrange for a peaky, raunchy kind of sound.
Let me again emphasise that quality of sound is highly subjective, so all the above examples describe starting points, not iron-clad rules. Nevertheless, these suggestions should definitely help get you started.
Interestingly, pitch transposers seem to be used most often for functions other than pitch transposition. Since the parallel harmonies produced by most inexpensive pitch transposers are not very useful musically (simply setting a major third, for example, will only sound right with certain notes), many musicians use these devices for doubling and chorusing, and even prefer them to many standard delays. But harmony synthesis has lots of applications, using effects unique to pitch transposers as well as some obtainable by other means.
1. Glissando.
This patch takes advantage of the regeneration control found on most pitch transposers. With the transposed pitch set very slightly higher than normal pitch, advancing the regeneration control recirculates each note and pitch shifts it, initiating a stepped, upward glissando effect (where each step interval is the difference between the interval of the transposed and original signals). Setting the transposed pitch slightly lower than the original signal produces a similar downward glissando.
Another excellent application for glissandos is with hand claps. By setting the unit for a subtle downward spiralling, the claps sound 'bigger than life' and much thicker than unprocessed hand claps.
2. 'Bell trees'.
Sustaining a single note with the regeneration advanced to nearly the point of feedback and then varying one of the pitch controls, produces a variety of useful effects, from pseudo 'bell trees' (seemingly endless upwards or downwards spiralling) to complex flanging-with-pitch-change effects. This effect again shows how adding regeneration can change a pitch-transposing device into an excellent special effects generator.
3. Expanding regeneration capabilities.
Processors with variable initial delay, or with patch points that let you insert a delay line to create a variable delay, greatly expand the limits of the regenerated sound. For example, by delaying the regeneration signal through an echo unit, you can slow down the speed of the glissando mentioned above. With a delay of, say, 0.5 seconds and a pitch setting a half-step above the original signal, hitting a note will result in an upwards glissando that goes up by one half-step every 0.5 seconds. One setting I particularly like is adjusting the pitch control for a signal that's a musical fifth above the original signal and adding some delayed regeneration.
When you hit a chord, you'll hear a fifth above that chord a fraction of a second later (depending on the amount of delay), then a fifth above the second chord, a fifth above the third chord, and so on until the regeneration dies out completely. Note that you only have to play one chord to initiate this chain of chords; the pitch transposing device takes care of the rest.
4. Polyphonic octave divider/multiplier.
Most octave divider boxes only work on one note at a time. However, by setting the interval control of a pitch transposer exactly one octave below the signal being processed, you have an octave divider that not only tracks your playing accurately, but also works with chords. Setting the interval control one octave above the signal being processed produces an octave multiplier effect.
5. Simulated string synthesis.
Setting the pitch an octave above the original signal (as in the previous example), but adding a slight amount of regeneration, makes for a sound that strongly emphasises high frequencies and synthesizes a bunch of upper harmonics. This creates an almost string synthesizer type of sound, even when added to an instrument like guitar; when applied to a true string synthesizer, the string simulation becomes even more lifelike. For best results, the pitch-transposed signal level should be set quite low; this is just to add a bit of sonic seasoning.
6. Flanging effects.
By setting the pitch control for a very slight amount of transposition (say, 1/10th to 1/5th of a semitone) and advancing the regeneration control, most pitch transposers can produce excellent flanging effects. Units with modulation sources or control voltage inputs can even produce the cyclic sweep associated with standard flangers; units without these extras will still give flanging effects, although the sweep effect will be somewhat different. For example, the sweep may start high, go low, splutter, and then repeat this process over and over. The exact way in which the device flanges will vary from model to model.
7. Speech compression.
Speech time-compression is a less obvious use of a pitch transposer, but one of great interest to salespeople, students, interviewers, and other people who need to process large amounts of spoken material rapidly. In this application, you can feed a microphone (or spoken programme material) into the device with the pitch set for one octave below the original signal, and then record this on a tape recorder running at, say, 7.5ips. By replaying the tape recorder at 15ips (exactly double the speed) the timbre of the speech returns to normal, but the rate of speech doubles.
Echo is commonly used to enhance an existing sound or add a feeling of 'ambience' (room or space acoustics), but you can also make an echo unit do a lot more than go 'echo-echo-echo-echo'.
1. Echo with added chorus effect.
Chorusing is a sound obtained by mixing a signal with its very slightly delayed twin, and subtly varying the delay time so that the twin acquires a more animated sound. With an echo unit, you can generate a twin signal every time an echo occurs. Subtly altering the amount of echo delay time via an LFO or other modulation source means that should the echo of, say, an A note appear when you play another A, you'll hear a choral effect. Consider another situation where a guitarist is playing a series of E chords, and generating echoes as well. These echoed E chords will often occur at the same time as the E chords the musician is playing, so we have a situation similar to chorusing in that there's a straight signal and its twin occurring very closely to each other with respect to time. Therefore, adding just a little bit of delay time modulation gives the slight amount of time-shifting necessary to create the familiar chorusing effect. Note, however, that this modulation must be tastefully applied. Too much modulation will create excessive pitch shifting, leading to an out-of-tune sound. Also, the balance/mix control should select a relatively equal blend of straight and delayed sounds.
2. Synchro-sonic echo.
This involves synchronising the echo rate to the tempo of a song by carefully adjusting the delay time. If you already know the song's tempo, simply divide 60,000 by the tempo to give the number of milliseconds per measure. If you don't know the exact tempo, here's an alternate way to figure out the proper delay setting.
Plug a microphone into the echo unit, turn the echo regeneration control way up (but short of runaway feedback), and generate a short 'click' or 'pop' sound while the song is playing. Since the regeneration control is up, you'll hear a long series of clicks. If the clicks occur at a faster rate than the tempo of the song, increase the delay time; if they occur at a slower rate, shorten the delay time. Eventually, you'll reach a point where the echoes seem to fall 'right in the pocket', although past a certain number of echoes, they may fall more and more out of sync with the tempo. If you need a sound with lots of regenerated echo, continue to tweak the delay time until the effect is just right. If you need only a tight, slapback kind of echo, then as long as the first six echoes or so match the song's tempo, you have no problem. Reduce the regeneration control to give one or two echoes, and you'll find they fall right on the beat.
3. Stereo separation.
At short echo times (20-80ms), you can spread a mono signal into stereo with the patch shown in Figure 7. With this patch, the balance control must be set for echo only; if any direct signal occurs at the echo unit output, the stereo image will shift away from centre (although this could be useful in some applications). Bear in mind, though, that an effect which sounds great live in stereo might not be so hot when mixed back onto tape in mono (the echo might be overly obvious, or at shorter delays, there could be phase cancellations/reinforcements), so monitor the results of your efforts in mono as well as stereo if you're working on a project that may be played back on a mono system.
Note that you can combine stereo separation techniques with synchro-sonic techniques to create echoes that appear in a different channel and follow a song's tempo.
4. Polyrhythms.
This simply requires patching two delay lines in series (delay 1's output feeds delay 2's input). Try setting the second unit's delay time for one-third the delay time of the first unit; that's just one of several interesting effects obtainable with this technique.
5. Using jacks on the back.
Some delay lines include an effects loop (patch points) on the back panel so that you can connect some other effect into the delay unit's feedback path. Although you can patch a wide variety of boxes in there, one of the most popular is equalisation. For example, by cutting the high frequency response with an equaliser, each echo becomes duller and more muffled - exactly the effect needed to simulate the sound of echoes decaying in a natural room environment. This also solves the problem of having later echoes 'step on' the source signal if the regeneration control is turned up, as progressively filtering out the highs means that later echoes do not just fade away in terms of level but also in terms of high frequency content.
Another variation is to filter out the low frequencies so that the echoes feed back only high frequencies. With vocals, this means that 's' sounds and other high frequencies are echoed to a greater extent than the rest of the vocal. The result is a bright, 'presency' sound.
6. Tape loop simulations.
Delays with echo times in excess of about two seconds or so can provide tape loop-type effects. Set the longest echo possible and turn the regeneration up to just short of feedback. As you play into the delay, it will layer your playing to create a rich, thick tapestry of sound. Many delays also allow for an 'infinite hold' setting where whatever is in memory can be 'frozen' for repeated playback. One useful application of this technique is to build up layers of sound through looping, then freeze the sound and improvise over the textured background [just like Robert Fripp -Ed.]
7. Hard reverb.
This is a fairly complex patch, but it gives a coarse, rude, 'grainy' reverb sound that you probably won't find as one of the preset programs in your preset digital reverb box. This requires two delay lines, along with a mixing console that features a stereo auxiliary bus (Figure 8). The object is to cross-pan the outputs and inputs of the two echo units. The aux sends add regeneration, so don't turn them up too far, or you'll create runaway regeneration effects.
Signal processing is a very open-ended field, because altering just one or two parameters can create an entirely different sound. As more and more processors start to allow real-time parameter control via MIDI, the options multiply even more. These days, signal processors are no longer static devices that sit in the signal path and change the sound. They can become an integral part of an entire music-making system. Some units even let you synchronise effects to MIDI timing, thus producing synchro-sonic echo effects without tedious calculations. And for onstage use, having programmable effects means that it is no longer necessary to reach down and change controls between songs or while playing a song. Simply programme all the sounds you'll be needing beforehand and use a MIDI footswitch to call up the desired programs as needed.
So go ahead, patch some signal processors into your setup and experiment. As you create sounds that no-one has heard before, you'll be developing your own unique 'sonic signature'.
Accidentally stranded on earth due to a bureaucratic mix-up involving a less-than-reputable intergalactic trucking firm, Craig Anderton has made the best of the situation by partaking in the unique earthly pleasure of playing music and editing America's premier music magazine, Electronic Musician.
© 1989 Electronic Musician magazine, 6400 Hollis Street, Suite 12, Emeryville, CA 94608, USA. Reprinted with the kind permission of the publishers.
An initial delay control sets the delay line's delay time. However, just leaving the delay at one setting produces a fairly static effect that is more like equalisation than a special effect. So, a modulation section sweeps the delay time in a cyclic manner to add some motion to the sound (like a tremolo or phase shifter). There should also be a depth (or width) control that lets you vary the range of this sweep from a subtle modulation to a wide range sweep that more closely resembles tape flanging.
Widen your horizons - FX tutorial (Part 1) |
Workbench - Signal Processors - the saga continues |
Special FX |
Delay Technology |
Off the Wall |
Fade away and radiate |
Compression Session (Part 1) |
Stereo Width Tricks |
De-Esser Project |
Workbench - Signal Processors — the saga continues |
Aural Examination - Aural Exciters |
Signal Processing With Sequencers |
Browse by Topic:
Feature by Craig Anderton
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!
New issues that have been donated or scanned for us this month.
All donations and support are gratefully appreciated - thank you.
Do you have any of these magazine issues?
If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!