The Ins and Outs of LFOs

Image placeholder title

String and woodwind players almost always enrich their long notes by using a technique called vibrato. To produce vibrato, string players wiggle the finger that's stopping the string so that the pitch moves up and down slightly during the course of the note. The result is a warmer, richer sound that stands out in a subtle way.

Image placeholder title

FIG. 1: Standard LFO waveshapes include sine, triangle, sawtooth, and square waves.

Because synthesizers were initially keyboard based, musicians couldn't add vibrato to synth sounds by wiggling their fingers. But vibrato was too useful a musical technique to go without, so synths were equipped with a mechanical form of vibrato. The pitch of a synth's oscillators could be modulated by a signal from a special low-frequency oscillator (LFO).

LFOs quickly proved useful for synthesis techniques in addition to vibrato. Today all synthesizers, most other electronic instruments, and many effects processors have LFOs. High-end instruments may give you four or more LFOs and let you route their outputs to a variety of synthesis parameters so that sounds can be animated in complex ways. In this column, I'll examine the features of the LFO and explain some of the things LFOs are used for.

Basic LFO Parameters

An LFO is a modulation source that produces a repeating signal. Except in a few special situations, the signal rises and falls in a regular, periodic way. Although the LFO is a type of oscillator, we can't hear its output. Instead, the output is used as a control signal for changing some parameter of a synth or an effect (for more on modulation, see “Square One: The Matrix” in the June 2006 issue of EM).

The most important parameters for programming an LFO are waveform, frequency, destination, and amplitude. The term rate is often used in place of frequency, and depth is used in place of amplitude.

Most LFOs have a set of basic waveforms, as shown in Fig. 1 (for more on oscillator waveforms, see “Square One: Making Waves” in the October 2005 issue of EM). If you think of an LFO as modulating pitch, it is easy to understand what these shapes do: the sine and triangle wave cause the pitch to rise and fall smoothly, for example, and the square wave causes the pitch to alternate between two values, which is a musical effect called a trill (see Web Clip 1). Your owner's manual probably has diagrams showing the waveforms that the LFOs on your instrument can produce.

The frequency of a low-frequency oscillator is, by definition, low. Some LFOs can produce signals having a frequency as low as 0.02 Hz (one cycle every 50 seconds). The upper frequency limit for many LFOs is around 20 Hz. Some LFOs can produce frequencies above 20 Hz, but when that type of signal is used for modulation, the result is a buzzy sound that covers up the individual cycles of the LFO waveform.

Image placeholder title

FIG. 2: The buttons along the top in the LFO section of Spectrasonics Stylus RMX may look like destination selectors, but in fact, RMX provides three separate LFOs, each with a single dedicated output routing. One LFO modulates the output amplitude of the voice, one modulates filter cutoff, and one modulates panning. Because RMX is a percussion instrument, it has no pitch LFO.

The LFO's destination is the sound parameter that it's modulating. If the destination is oscillator pitch, the LFO will produce vibrato or possibly a trill. If it's the synthesizer's output level, the LFO will produce tremolo. You'll find other destinations on many synths, including filter cutoff frequency, panning, and the amount of some timbral effect, such as FM. If no destination is chosen for the LFO, you won't hear it, no matter how you set the other parameters. Some LFOs are hardwired to specific destinations (see Fig. 2), but others can be connected to various destinations as needed.

Some synths have a switch that sets the LFO's destination in the LFO itself. You may be able to select only one destination for the LFO, or you may be able to switch on several destinations at once. Other instruments have a switch or a knob on the destination module for activating the LFO as an input.

If the LFO's amplitude is zero, again, no modulation will be heard. As the amplitude is turned up, the depth of the LFO's effect will increase. As with the choice of destination, some instruments let you control LFO amplitude in the LFO itself, while others have an LFO amount knob in the receiving module.

The amplitude amount parameter might be bidirectional, allowing it to be set to a value that is less than zero. That inverts the LFO waveform. In the case of sine, triangle, and square waves, inversion often makes little audible difference. But being able to invert the sawtooth wave is essential, so that it can produce either a rising slope or a falling slope. If the LFO's output can't be inverted with the amplitude knob or a dedicated invert switch, you'll probably see rising and falling sawtooth waves as separate choices in the waveform selector.

Take It up a Notch

Clever instrument designers have come up with a wealth of variations on the LFO. Here are a few of the features you might see:

LFO sync locks the frequency of the LFO to the tempo of a sequencer. The rate of a synced LFO is set in musically meaningful units, such as half notes or eighth-note triplets, rather than being expressed in abstract units, such as hertz. That is especially useful for creating filter sweeps that repeat every two or four measures, in time with your music.

The freerun/retrigger switch controls whether the LFO will restart its waveform each time you strike a key (retrigger) or ignore your keystrokes and continue to cycle (freerun). Retrigger mode is useful if you want a reliable modulation shape on each note, but it can sound a little artificial. Freerun mode sounds more natural.

The phase knob, which is active only in retrigger mode, controls the point at which the LFO waveform starts when a new key is struck. That is especially useful when the LFO is producing a trill; you can determine whether each note starts with the lower or the upper of the two pitches.

The LFO delay knob controls the amount of time that will elapse after the beginning of a new note before the LFO's amplitude rises from zero to its preset level. This onset delay allows you to play fast notes without using vibrato (which would obscure their pitch). The vibrato kicks in only on longer notes.

Some instruments have both an LFO delay knob and a slope or ramp knob. The purpose of this control is to smooth the transition between no LFO effect and the full LFO effect. Here again, the purpose of the parameter is to allow you to mimic the performance of a classical string or woodwind player, who will sometimes begin a long note with no vibrato and then add more vibrato during the course of the note, increasing the perceived intensity of the sound.

If the slope and delay parameters are bidirectional, you may be able to start each note with LFO modulation, and then have the modulation fade out. That can be used with a fast LFO frequency to add an interesting noise burst to the beginning of each note.

Most LFOs include an amplitude-modulation input in some form, which is often linked to Mod Wheel messages (CC 1). The purpose of the input is to allow you to control the depth of vibrato or some other type of LFO effect from the mod wheel while playing a keyboard part with your other hand.

LFO rate modulation allows the speed of the LFO to change during the course of a note. Increasing the rate of vibrato slightly as the depth increases makes the sound more expressive. Some LFOs can also change the rate slightly in a random way, which gives a less predictable and therefore more “human” sound.

Image placeholder title

FIG. 3: By clicking and dragging on the LFO waveform in Steinberg Xphraze, you can change its shape. The smoothing knob controls how smooth or stepped the waveform will be.

LFOs on software synthesizers increasingly have more waveforms than the standard sine, triangle, square, and sawtooth types. You may be able to design your own waveform (see Fig. 3), import a complex sampled wave, or choose noise (a random, nonperiodic signal) in place of a repeating waveform.

Another option often found among the waveform selections is sample-and-hold. This effect goes back to the early days of synthesis. The LFO module produces a stepped signal in which the level of each step is random rather than a repeating waveform. The speed with which new steps are produced is controlled by the LFO's rate parameter. Sample-and-hold is typically used to modulate pitch or filter cutoff; it's a special effect that sounds very electronic.

Spice of Life

Not all synthesizer patches need the animation provided by LFOs. But if there were no such thing as an LFO, synthesizers would sound so dull and bland that no one would want to listen to them. The LFO is as essential an ingredient in sound design as spices are in cooking.

Jim Aikin writes regularly for EM and other publications and Web sites. He is the series editor of Backbeat Books' Power Tools music-technology books.