You are here

Using Effects With Keyboards

Tips & Techniques By Paul White
Published July 1994

Most of us can set up a suitable vocal reverb treatment, but what's the best way to deal with all those synthesized and sampled instruments? Paul White offers a few suggestions.

With such a bewildering array of effects offered by the current crop of stand‑alone processors and workstations, choosing the best effects processing isn't always easy. In the case of a keyboard workstation, it's tempting to use whatever effects are already programmed in, but that neglects the true creative potential of the machine. The same is true of stand‑alone effects units, where the factory presets often seem to provide more than enough choice. However, even more important than the range of effects on offer is the need to match sounds with these effects, so that the end result has both purpose and musical relevance.

If you've done any home recording, the chances are that you'll be able to come up with sympathetic reverb treatments for drums and vocals, but what, if anything, should you do to synthesized and sampled sounds? Because of the creative nature of music, there are no inviolable rules, but in many cases you might find it helpful to consider the acoustic surroundings in which the original instrument could have been played. Even in the case of a completely synthetic sound, you can often imagine the environment in which you'd like it to be heard, and go some way towards creating that. The aim of this article isn't to lay down hard and fast rules, but to try to establish a few practical guidelines which you can then build upon, based on natural acoustic principles.


Virtually every synth built over the last five years includes some form of choir sound or vocal pad, but in their raw state, they often don't sound much like the real thing. I've found that the best results are achieved by layering two or more choir patches from different instruments, and then adding effects.

Real choirs are generally associated with churches, cathedrals or other large buildings, which implies longish reverb times. But you can do more than simply slap on a five‑second reverb. Going back to natural acoustics, choirs are comprised of many people, all singing slightly out of tune with each other, and at slightly different times. If the choir is a good one, these human variations will be small, but they can never be eliminated. A good first step to simulating this is to use a stereo pitch‑shifter/delay algorithm, to create two detuned and slightly delayed copies of the original sound. A detune setting of between five and ten cents is adequate to produce a natural chorus effect, and if one output is panned left and tuned down slightly, while the other is panned right and tuned up slightly, the nominal pitch will remain the same, and the stereo spread will be enhanced. To simulate the timing delays of the different singers, the two detuned signals can be delayed by different amounts between 20 and 50ms. If more than one synth or sampler is being used to create the basic choir sound, these too may be detuned slightly. If there are only two layers of sound, one should be tuned slightly flat and the other slightly sharp, so as to maintain an average pitch that is still in tune.

Now we can start adding reverb, and because cathedrals tend to include a lot of hard surfaces, a bright hall setting (with the early reflections level, if you have one, turned right up) should bring you somewhere close. To reinforce the illusion of distance as well as space, try a pre‑delay time of between 50 and 100ms. An alternative to this is to delay just one of the reverb outputs by 50ms or so, which will create a sense of left/right movement. This is less natural than pre‑delay, but is a very pleasing effect in its own right.


Like choirs, string sections are made up of many performers, all playing slightly differently. You can use the same treatments as you would for a choir, but because the usual venue is a concert hall, the appropriate reverb treatment should be around three seconds, rather than five. Gentle stereo chorus may be used, either instead of or as well as pitch detuning, and this should be applied before adding the reverb. To prevent the final sound becoming too muddy, use the reverb sparingly, especially if the string patch has a slowish release time. If you have separate control over the early reflections level, you can increase this, to reinforce the illusion of many people playing together.

When layering string patches, try to pick sounds with slightly different attack rates, to create the effect of the string sound building up. One trick I've used with both string and choir sounds is to use a second layer an octave higher than the first, and with a noticeably slower attack. Analogue string pads can also be combined with digital string pads or samples to good effect.


Brass instruments tend to have an obvious attack when played hard, and in an ensemble, small timing differences become quite obvious. A stereo multitapped delay, with randomly set timings between 20 and 70ms, works well to simulate this effect, and, to avoid blurring the sound too much, a short, bright reverb treatment often gives the best result. A two‑second plate tends to work well in a pop or rock context, though for a classical sound, a good concert hall patch set to decay for between two and three seconds is ideal.

Brass ensemble sounds may be further thickened by the use of pitch‑shifter, detune, or chorus, as described in the Choir and String sections.


Though part of the standard orchestra repertoire, flutes are often used in contemporary work, where they are frequently treated using long reverbs, echoes, or combinations of both. The same is true of pan pipes, shakuhachis, or indeed any wind instruments which operate along the same lines (air being blown over an opening). As a general principle, long reverb or echoes work best on sparsely orchestrated pieces, as can be confirmed by listening to a selection of New Age compositions, but in an orchestral or ensemble context, it may be safer to err on the side of more natural acoustics. Concert hall patches are quite satisfactory, though the more intimate sound of a 'tiled room' or 'medium room' patch helps draw the listener into the music.

Because of the tonal purity of flutes and their ethnic cousins, detuning and chorus effects tend to detract from the character of these instruments, and are best omitted, unless used very sparingly.

Bass Sounds

The low end of the audio spectrum can easily become confused and cluttered if treated using long echo or reverb, which is why most bass sounds tend to be left fairly dry. Short delays may be used to create automatic double‑tracking (ADT) or doubling effects, or you can try gated reverbs and early reflection patterns to create stereo spread and space without clogging up the mix. Effects such as flanging can also be effective on electronic bass sounds, because they add interest and movement without 'smearing' the sound. You can treat fretless bass slightly more adventurously, and in slow, sparsely orchestrated music, try combining both chorus and reverb to create a warm, sensuous feel.

Where there is a need to create a greater sense of bass energy, compression or limiting can be used to increase the average sound level without increasing the peak level. Setting the compressor attack time to between 10 and 50ms can help emphasise the attack of percussive bass sounds. On a practical note, bass sounds are normally panned to the centre of the mix, so that the low‑frequency load is shared by both loudspeakers rather than only one. This helps create a louder‑sounding and more stable mix.

Pad Sounds

Because pad sounds tend to be ongoing, there's little point in adding echo or reverb, because all the gaps in the mix are already full. In most instances, if you need to add interest, try using gentle chorus or flanging, ideally in stereo, to create a sense of width and movement. The human hearing system soon learns to ignore repetitive or constant events such as the tick of a clock, or the whirr of a fan heater, and the best way to keep the brain interested is to introduce change. Detuning patches are also effective in widening pad sounds.

Another way to create movement is to actually move the sound! If your effects unit includes a panner, try moving the sound from left to right and back, at a speed related to a multiple of the tempo of the song.

A Final Word

Ultimately, effects are tools to create an illusion of some kind, and every illusion starts with a good picture — in this case, your mix. If a mix can't stand on its own, without effects, the chances are that it won't improve all that much when the effects are added. On the other hand, get the basics right, and the right effects will almost suggest themselves.

The ART Of Restraint

If you listen to a selection of mixes from respected producers, you may be surprised at the apparently limited use of effects. This is because a good producer knows when to leave an effect out as well as when to put one in. Vocals will be treated with reverb, but not to the extent that they are rendered unintelligible or pushed back in the mix, and the rhythm section will usually be tight and crisp, with plenty of space. Pads are mixed well back, so as not to conflict with the main melody or vocal line, and such effects as are used are applied only after consideration of what instrument is playing, what else is playing at the same time, and how much space there is left in the mix to work with. A useful tip here is that stereo reverb doesn't always have to be used in stereo. If you want to pinpoint a sound in a mix, pan the reverb to the same point as the original sound, or to create more movement, put the dry sound over at one side of the mix and all the reverb over at the other!