Polyphony

Polyphony is the property of an electronic musical instrument which describes how many notes it can sound at one time. An instrument which can produce multiple notes at a time is said to be polyphonic. If an instrument is capable of playing 16 notes at once, for example, it is said to have 16-voice polyphony or 16-note polyphony. By extension, polyphony may be used to describe traditional instruments. For example, a piano may be said to be polyphonic while a flute is not. Instruments which are not capable of polyphony are monophonic. The property of polyphony may extend to devices not generally considered to be musical instruments. Examples include mobile phones with polyphonic ringtones and video game consoles.

Polyphony was originally used to describe the ability of synthesizers in particular to play multiple notes at a time. Early synthesizers were monophonic. While synthesizers were revolutionary in many respects, a monophonic keyboard instrument was quite limited. Serious multi-voiced recordings created with monophonic synthesizers such as the Moog modular synthesizer were tedious to create. One such landmark recording was Switched-On Bach, a collection of pieces by Johann Sebastian Bach arranged for synthesizer and performed by Walter Carlos. Such a recording would have been considerably easier to make with a polyphonic synthesizer.

Synthesizers traditionally use a voltage-controlled oscillator to generate an electric signal which, when amplified, makes a sound. A monophonic synthesizer connects the keys of a keyboard to an oscillator (or a pair of oscillators working in unison). Whenever a key is pressed, the oscillator oscillates at a specific rate producing the pitch corresponding to the pressed key. When another key is pressed, a different note is sounded by the oscillator. This was a relatively easy thing to accomplish using analog electronics. However, multiple oscillators working independently are a considerable challenge to implement. To double the polyphony requires more than double the amount of circuitry because, not only are the number of oscillators doubled, but the electronics also need to function as a switch, connecting keys to free oscillators instantaneously. The switching function also implements an algorithm which decides which notes to turn off if the maximum number of notes are already sounding when an additional key is pressed. There are several ways to "decide" what happens when this occurs:


 * Turn off the first note sounded and use the newly freed oscillator to play the new note
 * Turn off the last note sounded and use the newly freed oscillator to play the new note
 * Ignore the newly depressed note

Modern synthesizers and samplers use additional, multiple, and/or user-configurable criteria to decide which notes are turned off. The algorithms which make these decisions are designed to create the impression of extended or unlimited polyphony and are quite sophisticated. They are only feasible through the use of digital technology.

The earliest polyphonic synthesizers arrived on the scene in the early 1970s, but viable polyphonic products were not available until the late 1970s and early 1980s. One notable, early polyphonic synthesizer was the Roland Jupiter-4, which was released in 1978 and had four-voice polyphony. Six-voice polyphony was standard by the middle 1980s. With the advent of digital synthesizers, 16-voice polyphony became standard by the late 1980s. 64-voice polyphony was common by the middle 1990s. Today, a music workstation such as the Korg Triton Extreme has 120-note polyphony.

Extended polyphony may lead one to wonder why it is necessary for an instrument to play 16 or more notes simultaneously when a musician can depress, at most, 10 keys at a time. There are several reasons:


 * Even with ten fingers, it is possible to play more than ten notes at once.

Notes may continue to sound even after a key is released. The synthesizer's resources may still be in use to produce the sound of the previously struck notes tapering off. Use of a sustain pedal exacerbates the situation. A realistic keyboard performance which includes chords, a melody, and a bass line is not really possible with limited polyphony.


 * A "sound" may be generated by more than one oscillator.

On many synthesizers, multiple oscillators can be assigned to create a timbre, also called a "sound" or a "patch". Naturally, the use of multiple oscillators allows more complicated sounds to be produced. If a synthesizer has, say, 16 oscillators, it may be capable of 16-note polyphony, but only when simple, single-oscillator sounds are produced. If a particular patch requires four oscillators, then the synthesizer is only capable of 4-note polyphony. Of course, such an instrument is still marketed as having 16-voice polyphony.


 * Synthesizers may be configured to layer multiple timbres.

On many synthesizers, two or more patches may be layered. A synthesizer with the capability to play multiple timbres at the same time is said to be multitimbral. For example, a musician may wish to play a pipe organ sound and a string ensemble sound at the same time. Assuming the pipe organ patch requires two oscillators and the string patch requires two oscillators, pressing a single key requires four oscillators to engage at once. Thus, additional polyphony is beneficial when layering sounds on a multitimbral instrument. Multitimbral instruments are always polyphonic, but polyphonic instruments are not necessarily multitimbral. Some multitimbral instruments have a feature which allows the user to specify the amount of polyphony reserved or allowed for each timbre.