Modulation Effects

Do your productions lack texture and movement? Your DAW’s modulation plug‑ins might offer a solution.

Every DAW includes a host of modulation plug‑ins, and there are plenty of affordable hardware options too. They all have in common that the effect is created by one parameter being modified over time by a control signal, or ‘modulation source’. The most common of these is the low‑frequency oscillator (LFO), which always produces a somewhat cyclic result whatever its waveform, but it could be the input signal amplitude, an ADSR‑style envelope generator, an expression pedal, MIDI data or CV signal. It’s sometimes possible to combine modulation sources for more complex control signals, or to modulate modulation sources. And, of course, you can use DAW automation to modulate almost any plug‑in parameter very precisely.

In this article, I’ll touch on some other types but my main focus will be on the various conventional modulation effects — you might be surprised by what you can achieve with them! I’ll explore what each main type does, what’s going on ‘under the hood’, and what they might bring to your productions. Most are LFO‑controlled, and the most commonly used LFO wave shapes are sine and triangle, which provide the smoothest‑sounding result. But you’ll often find square waves used too, and even random waveforms or random steps in some cases.

Tremolo

The earliest electronic modulation effect I know of is tremolo. Unlike the tremolo arm of a guitar (misnamed by Leo Fender, this actually produces vibrato!), the tremolo effect is a modulation of signal level. It has been built into many guitar amps and pedals and produces a pleasing ‘undulation’ at low modulation depths, but a stronger pulsing or even a ‘chopping’ effect when the depth is greater. In the DAW world, tempo‑sync’ing the modulation rate can be incredibly useful, as it makes it easy to use tremolo to reinforce or contribute to the rhythmic feel of the track. That said, you could leave the tempo loose for that old‑school surf guitar sound.

In most modulation effects, a low‑frequency oscillator (LFO) changes the behaviour of one or more parameters. Simple sine, saw and square waveforms can take you a long way, but some plug‑ins offer more options. For example, Soundtoys’ Tremolator allows you to edit the wave shape, and includes presets for the ‘imperfect but charming’ LFO waveforms of various vintage analogue devices.

A sine‑wave modulation produces the familiar pulsing effect, whereas a square wave provides that more strongly‑defined chopping character, and can be exploited to turn a static pad or long reverb wash into something rhythmically useful. Some plug‑ins allow you to offset the modulation waveform in time (often by adjusting a ‘phase’ control). This means that you don’t necessarily have to have the loudest sound sitting right on the beat. And where a deep tremolo’s LFO is configured in opposite polarity on the left and right channels of a dual‑mono signal it will create an auto‑pan effect.

Electric guitar is an obvious candidate but you can use tremolo on pretty much any sound source you like. I’d not generally use it on vocals except as a special effect (Kate Bush’s ‘Waking The Witch’ is a striking example). Less obvious applications include using a very slow LFO rate to vary the blend of two sounds playing simultaneously (eg. layered pad textures). To achieve this, put an instance of the same tremolo plug‑in on both tracks and, as with the auto‑pan effect, flip the LFO phase or polarity on one instance, so one track is at its loudest when the other is at its quietest. If you don’t want the changeover between parts to be complete, reduce the depth of the tremolo — both parts will always be audible but their relative levels will change. This can be great for creating subtly shifting textures that keep the listener’s interest without too obviously drawing attention.

Vibrato

Most of us will be familiar with vibrato as an effect controlled by the mod wheel of a synth, a guitar’s so‑called ‘tremolo arm’ or a string player’s fingers, but there’s a place for vibrato as an insert effect too. The modulation mechanics are similar to those of tremolo but the LFO modulates pitch, not volume. At the heart of this effect is a very short delay line, and modulation of the delay time creates the familiar effect. Thus, the output is delayed slightly relative to the input but this delay is too short to be noticeable. Again, for the most ‘natural’ sounds choose a sine or triangle LFO shape, while square wave will produce a pitch‑trilling effect.

Used subtly, gentle pitch-wavering can add a delicate feel to pad or lead sounds, but take it too far and you end up with a deep, churning sci‑fi warble. In most cases that will sound utterly unmusical, but more extreme square‑wave‑controlled settings will often throw up happy accidents when experimenting with drum loops; hi‑hat patterns and occasional cymbal hits can become very interesting if you sync the effect to the loop tempo. Slower pitch modulation can also be used to emulate the worn mechanics of a tape echo machine or cassette deck, and plug‑ins often combine this with other processes (noise, saturation, filtering etc) that degrade the sound. Similarly, slow pitch ‘wow’ can be used to imitate warped vinyl.

Used subtly, gentle pitch-wavering can add a delicate feel to pad or lead sounds, but take it too far and you end up with a deep, churning sci‑fi warble.

Chorus & Ensemble

The next two effects, chorus and flanging, are essentially spin‑offs from vibrato: both rely on mixing the dry signal with a slightly delayed version of the pitch‑modulated signal. Choruses are designed to create the illusion of two or more instruments playing the same part at the same time — there are always lots of small tuning and timing differences between any two human performances, no matter how brilliant the players. The chorus effect’s LFO‑driven approximation of the same thing might be more obviously cyclic, but it became very popular in its own right and was used extensively in the 1970s, through the ’80s and into the early ’90s: you can hear it everywhere from pop, new wave, shoe‑gazing and glam rock to grunge, a famous example being the main guitar on Nirvana’s ‘Come As You Are’. It fell out of fashion, with lots of musicians feeling that it had become a cliché, but it seems to have enjoyed a renaissance in the last few years.

Given those examples, it’s no surprise that chorus can work well with electric guitar, but one variation I’ve used both live (with two amps) and in the DAW is to hard‑pan a dry signal to one side and send it to a chorus that’s hard‑panned to the other. This gives a really nice wide chorus effect, without losing too much definition.

Roland’s CE‑1 and Dimension D (modelled here in Universal Audio’s plug‑ins) built on the chorus effect to create a smoother, lusher‑sounding ensemble effect. Most DAWs will offer something similar, though the imperfections in these vintage effects mean they have a charm of their own!

Another early application was in creating plausible string sounds from keyboards in the pre‑sampler era. In these so‑called ‘string machines’, a basic waveform was processed using one or more stages of chorusing to create a string ensemble effect. (A similar effect can be created in synths by applying pulse width modulation to a square wave, or layering two oscillators, one of them slightly detuned.) To create a more believable ensemble effect, Boss first developed their CE‑1 in 1976, based on the chorus/vibrato circuit found in the Roland JC-120 amp, and a few years later Roland released the now‑legendary SDD‑320 Dimension D processor, which could be controlled with combinations of just four buttons. Then, in the mid‑’80s, Yamaha’s SPX90 processor included a Symphonic patch, which proved very popular. Such ‘ensemble’ effects were based on multiple subtle chorus effects running at the same time but at different modulation rates — this helped to disguise the cyclic nature of the more basic chorus pedal. Most DAWs now come bundled with both traditional chorus and dedicated ensemble plug‑ins, but the latter can be imitated by using a few chorus instances in series.

Chorus can be a really effective way to add texture to keyboard parts, and is particularly popular for Rhodes‑style electric pianos. But it can work wonders on bass parts, including fretless bass, too. (A good approach can be to have a mono bass part and a stereo chorus effect, the wet signal being high‑pass filtered to create stereo interest without robbing the part of low‑end power.) An ensemble effect may also be used to thicken and diffuse backing vocals or to add richness to string or brass parts.

Here’s perhaps a less obvious way to use chorus: you can often add more texture to a sound by using it on an effects send, with reverb or delay. The result will be more subtle if the chorus is placed before the reverb, but a post‑reverb chorus can be musically valid and either can add a richness to the reverb that doesn’t sound too obvious. You can try the same thing with a vibrato effect or a flanger.

Flanging

Although they’re similar in architecture to the chorus, flangers tend to use shorter delay times and often incorporate an adjustable feedback path. If a sound is added to a very slightly delayed version of itself, some frequencies combine additively whereas other cancel out, and this results in what we call a ‘comb filter’ — a filter with a lot of narrow peaks and troughs. When the delay time is modulated, the frequencies that are ‘picked out’ by this comb filter change, and this is what gives rise to that familiar ‘whooshing’ sound. Increase the amount of feedback and that whoosh is emphasised, eventually turning into a very strong whine.

A flanger modulates the playback delay time for a duplicate of the source. When the source and delayed version are combined, a comb filter is created, and this moves up/down the spectrum as the delay changes, causing the familiar ‘whoosh’ sound.Flanging can create a very psychedelic trippy sound on vocals but it can work its magic on anything that has a rich harmonic content such as distorted guitar or synth parts. It saw a lot of use back in the day, perhaps most famously on the drum fills of ‘Itchycoo Park’ by the Small Faces, though back then it was created in the old‑school way: running the same sounds on two tape recorders, started at the same time and with their output signals mixed together, and one machine forced to change speed slightly, either by using a varispeed control or by applying finger pressure to the tape reel flange. An interesting characteristic of genuine tape flanging is that modulating the speed of two tape machines can result in a ‘delay’ moving before the dry signal. To emulate this ‘through‑zero’ flanging in a plug‑in or stompbox flanger requires a delay, so many don’t. By way of example, Apogee’s Clearmountain’s Phases plug‑in does include this effect, alongside the modelled MXR hardware devices.

Overt flanging can seem like overkill, due to its popularity in days gone by, so it tends to be used more sparingly or subtly in modern productions. Using a flanger to affect only a reverb or delay send can be a good approach: put the flanger before a reverb so the reverb will smear the effect, such that it adds a useful texture to the reverb without the flanger’s sweeping sound being obvious.

Note that many devices and plug‑ins allow you to switch off the flanger’s internal LFO modulation and adjust the delay time to produce a static comb‑filtering effect. This will produce a metallic ringing tone, especially at higher feedback settings, and can be a useful treatment for some rhythm parts, such as drum loop ‘toppers’, and was a feature of the original EHX Electric Mistress Flanger. And, of course, you can also use external modulation sources or DAW automation to make it move in ways the stock LFO can’t.

Phasing & The Uni‑Vibe

The phaser could be thought of as the flanger’s more polite sibling, and you can dig much deeper into this effect in Hugh Robjohns’ excellent 'How Phasers Work' SOS August 2021 article, but I’ll give a quick overview here. Again, phasing works by combining a dry signal with a processed one, but rather than the flanger’s delay the phaser’s wet signal passes through a number of phase‑shifting stages — usually between three and six, though some designs allow more. This again results in selective frequency cancellation and addition, but there are fewer peaks and troughs than in a flanger’s comb filter. Phasing started life as a guitar effect and remains popular in the stompbox market, but it can also sound very effective on a variety of sound sources, such as drum loop toppers, synth pads, pianos and so on.

The Uni‑Vibe offered a particularly smooth, watery‑sounding variation on the phaser effect. Several plug‑ins now model this and some, including Blue Cat Audio’s Polyvibe, offer a little more control than the original hardware.

One of the most famous (and earliest) types of phaser is the Uni‑Vibe, a pedal whose effect is controlled by four photo‑resistors placed around a pulsating light source. A distinctive sound is produced by the combination of the light‑source modulation characteristics and the four phase‑shift stages all being slightly different. Initially an attempt to approximate the sound of a rotary speaker, it failed in that aim but this smooth, watery‑sounding effect soon carved out its own niche. You can hear it on some of Jimi Hendrix’s gentler songs. David Gilmour and Robin Trower made good use of it too, and lots more artists have followed in their footsteps, including Pearl Jam, Smashing Pumpkins, and Queens Of The Stone Age. Original models fetch astronomical prices, but there are various hardware and software emulations and ‘evolutions’ of it.

An interesting variation that can be applied either to a phaser or a flanger is often termed a ‘barber’s pole’. Loosely inspired by the Shepard Tone illusion, the idea is to create the impression of perpetual motion in one direction, with the filtering effect always moving up or always down in pitch, rather than the usual alternation between up and down. The effect can be created in a number of ways, including crossfading two instances of the effect under LFO control. In the plug‑in world, iZotope’s Möbius Filter is designed to produce such an effect, as are some of Sinevibes’ plug‑ins, among others.

Rotary Speakers

The rotary speaker, exemplified by the various Leslie cabinets designed for use with tonewheel organs, are arguably the most complex of the ‘standard’ modulation effects, in terms of what they actually do to the sound. In these speakers, the low frequencies are directed by a rotating drum, with an opening at one side and an in‑built angled reflector, while the high frequencies are directed via a rotating horn, balanced by a dummy‑horn counterweight. The drum and horn run at different speeds, controlled by motors and pulleys, and when the speed is changed from low to high or vice versa, the inertia of the moving parts means not only that the speed change isn’t instant but also that it happens at a different rate for the high and low frequencies. Add to this the imperfect response of the combined high‑ and low‑frequency speaker sections, internal reflections within the cabinet, and the distortions in the amplifier, and you get a unique sound that combines Doppler shift, amplitude modulation and complex filtering. There are also some less complex rotary speaker designs which don’t use separate high and low drivers — for example, the Vibratone or the Gibson Maestro Rover RO‑1 — and while these generally deliver a less complex effect, they can still sound very musical.

Being unbound by the limitations of physical mechanics, some plug‑ins, such as Klevgrand’s Spinn, offer interesting variations on the rotary speaker effect.If the rotary speaker had only been used with organs it might not have ended up being quite so relevant but, as ever, guitarists (notably George Harrison and Dave Gilmour) started playing with them and the effect became more popular. Producer Alan Parsons also used a Leslie to treat some of the backing vocals on Pink Floyd’s Dark Side Of The Moon and, again, plenty more artists and engineers have since followed in those footsteps. Of course, you can try it on anything, and in modern productions you could generally look to a slow rotary speaker emulation as an interesting alternative to a chorus or phaser effect, including to process a reverb or delay aux channel. All three effects produce different forms of gentle modulation, and if you have a good Leslie 122 emulation (arguably the most revered rotary speaker model), the tonal attributes will be be very different from those of the original sound due to the modelling of the driver characteristics and the effect of the cabinet.

The fast rotary effect, as heard on the guitar in the middle section of Cream’s single ‘Badge’ (allegedly played by George Harrison) and on ‘Echoes’ by Pink Floyd, sounds much more obvious. In fact, whatever you pass through it is treated to a rapid trilling sound. It can be used on any keyboard sound, though a fast rotary setting does tend to lend an organ‑like quality to most keyboard sounds so while it can be a lovely effect, it’s probably not one you’d want to overuse. In fact, it’s often best kept as an occasional effect, with the transition from slow to fast and back again injecting an interesting sense of movement.

An authentic Leslie‑type speaker is complex, and thus difficult to emulate accurately, but plenty of plug‑ins offer a reasonable approximation and several do an excellent job. So too do some pedals, the most notable versions coming from Neo Instruments and Strymon. Some plug‑ins offer more flexibility than the originals, of course, and if you want the rotary effect without excessive coloration, Eventide’s Rotary plug‑in works extremely well; it models the rotary behaviour without the actual amp, so you can have a full‑range sound or choose your own flavour of filth! Klevgrand’s Spinn is an interesting plug‑in which splits the audio into four frequency bands and then provides a separate rotary emulation for each band, and again that offers more options.

Filters & Wah

Any filter’s parameters can be modulated too, of course, and for this both LFO and envelope control sources are common. The familiar wah‑wah pedal is a modulated band‑pass filter effect, controlled manually by a pedal rather than an electronic source, and where you have control over the bandwidth, you’ll find that a low‑Q band‑pass filter gives you a richer, fuller vowel sound. You don’t have far to look to find automated examples though. The envelope filter of a synth works in much the same way, and there are dedicated pedals such as the EHX Q‑Tron which can control the filter from the input signal’s envelope, to produce a variety of effects depending on the filter settings, modulation direction and playing dynamics. Others offer LFO control too.

A fast auto‑wah type effect can be useful for funk rhythm guitar or bass parts, while a slower filter frequency rise combined with less resonance can add a soft‑attack effect to notes or chords — something exploited by Mike Oldfield on the Songs Of Distant Earth album. Alternatively, modulating the filter from an LFO can produce a useful alternative to phasing or rotary speaker effects, and some pedals offer this option too — it’s also very easy to set up in your DAW. It’s also worth experimenting with modulated filters used in parallel with the dry source.

Other Modulation Effects

I’ve focused mainly on effects that can be modulated by envelopes, LFOs, step sequencers and so on, but some modulation effects work very differently. For example, you’ll usually find a ring modulator in the modulation section of your DAW’s plug‑in folder. This takes two inputs and then outputs their sum and difference frequencies but without any of the fundamentals being present. When fed with two synth tones at suitable musical intervals the results can be fairly musical, but with non‑related sources the sound can seem alien and very unmusical. If one of the inputs is fed from a fixed‑frequency sine‑wave oscillator at 50‑100 Hz and the other input is a voice, the familiar robotic voice beloved by Dr Who fans everywhere will issue forth! Sometimes, the more wayward sound of a ring modulator can be mixed with some of the dry sound to add interesting harmonics and overtones, something occasionally exploited by guitarist Jeff Beck.

Another less typical type of modulation is the Vocoder, in which the spectrum of one input (the modulator input) is imposed as a complex filter on the other, known as the carrier input. This effect is often exploited by taking the spectrum of a voice and imposing that onto a keyboard sound fed into the ‘carrier’ input to produce the very familiar ‘talking keyboard’ sound. A low‑tech equivalent of the vocoder is the talkbox — a guitar or keyboard is fed into a small amplifier and speaker, then the sound is routed through a tube into the performer’s mouth. By mouthing vowel shapes or words, the instrument can be made to appear to talk — the end result is picked up using a vocal mic.