Monophonic vs. Polyphonic Synths Explained

A monophonic synthesizer plays one note at a time, a polyphonic one plays several notes that each get their own sound shaping, and a paraphonic one plays several notes that share a single filter and envelope. If you want chords and pads you need polyphony; for bass and lead, mono is often not just enough but the better and cheaper choice. How many voices you need depends less on the number of fingers you use than on how long your notes ring out.

What a voice is, and why oscillators do not count

A voice is the complete signal path for one note: oscillator(s), filter, amplifier and their envelopes. A synth's polyphony is the number of these paths that can run at once. The number of oscillators tells you nothing about it: a Minimoog has three oscillators and still plays one note, because all three feed the same filter and envelope. What defines polyphony is the number of complete voice engines. On digital instruments and software that number scales easily; on analog hardware every extra voice means real components, board space and calibration effort.

For the technical background on oscillators and filters, see polyphony, oscillators and filters.

Monophonic, paraphonic and polyphonic compared

Monophonic

If you play a second note while holding a key, the first one cuts off or, with glide switched on, slides into the second. Since only one signal path is needed, monophonic instruments are simpler and usually much cheaper to build. Many classics are mono on purpose: the Minimoog (built 1970 to 1981, roughly 12,000 units, three oscillators, a 24 dB ladder filter), the TB-303 or the MS-20. A single focused voice suits bass and lead, and designers of mono instruments can concentrate entirely on sound quality and layout.

Paraphonic

Paraphonic synths are the middle ground: each oscillator may respond to its own key, but all of them run through one common filter and envelope. Typical examples are the Korg Mono/Poly, the ARP Odyssey (which only manages two notes, often called duophonic) and the Moog Subharmonicon. The Korg Poly-800 has eight oscillators but only one filter and one envelope. You can hear the consequences:

• The envelope cannot restart per note. Depending on the instrument, playing a second note over a held one either does not retrigger at all, or retriggers every held note along with it.
• Chords sound as one block: all notes open and close the filter together, so there is no independent articulation of individual notes.
• The tone is often idiosyncratic, and many players like it for exactly that reason.

The word itself has a changing history: a manufacturer introduced it in 1977 as a marketing term for a guitar-synth system with a slightly different meaning. Today it almost always means the shared-signal-path variant.

Polyphonic

On a truly polyphonic instrument every voice has its own filter and envelope, so one note can ring on while the next hits with a sharp attack. The voices all sound alike, since they share the same panel settings. The first truly polyphonic synthesizers appeared in the mid-1970s (Yamaha GX-1, CS-80, Oberheim Four-Voice). The five-voice Sequential Prophet-5 of 1978 was the first fully programmable polyphonic synthesizer, one that could store every setting.

How many voices do you need?

The number of keys held at once is only half the calculation, because fading notes occupy voices too. As long as the release stage is running, the voice is busy.

A common recommendation in buying guides is to plan for at least four to eight voices for chords and pads. As a practical rule for sustain-heavy chord playing, take the largest number of notes you hold at once and double it, so earlier chords can ring out without being chopped off by new ones. These are rough values and depend on the sound and your playing.

History shows many players got by with little: the Juno-106 (1984 to 1988) has six voices, the Prophet-5 five, the Jupiter-8 (1981) eight. The Yamaha DX7 (1983) offered 16 digital voices. Modern digital instruments and software often provide 64, 128 or more; the Yamaha Montage offers over 100.

When you run short of voices

1. Shorten the release of your pads. This is the most effective step, since fading notes otherwise block voices.
2. Switch off unison when playing chords (see below): it eats voices.
3. Voice chords in open positions without doubling notes across several octaves.
4. Use the sustain pedal sparingly, since a held pedal keeps every note open.
5. On software and digital instruments, raise the maximum voice count where it is adjustable.

Voice stealing

If you play more notes than there are voices, the synth has to take a voice away from a sounding note: the voice is stolen. Usually the older note is cut off instead of fading out naturally. This is normal behavior for any polyphonic instrument at its limit, not a defect.

Which voice gets hit is decided by the algorithm. In practice you meet these strategies:

• The oldest note is stolen.
• Notes already in release are preferred, because that is the least audible.
• The lowest or highest note is sacrificed.
• The most recent note is stolen, so a held bass note in your left hand stays while chords and melodies run on the right.

Which variant a unit uses is usually documented in its manual and differs by model. An audible click when stealing points to a simple implementation; good designs fade the old voice out briefly. The more voices an instrument has, the less often and the less obviously voice stealing occurs.

Note priority on monophonic synthesizers

If you press several keys on a mono synth, it has to decide which note to play. Note priority governs that:

• Low (lowest note): the lowest key held wins. The Minimoog works this way. Handy for basslines, because your left hand holds low notes while whatever you play above does no harm.
• High (highest note): the highest key wins. Good for trills and legato melodies played from the upper hand.
• Last (latest note): the most recently played note sounds, higher or lower. This is the most intuitive option for lead playing.

When you release a key, the synth jumps back to a still-held note, depending on the priority mode. Some instruments offer all three modes, like the Memorymoog with low, latest or high priority and a choice of single or multiple triggering. The trigger setting decides whether the envelope restarts for each new note (multi trigger, punchy) or runs on through overlapping notes (single trigger, legato). For flowing leads pick single trigger, for staccato basslines multi trigger.

Unison, layer, split and multitimbrality

Unison

In unison mode several voices play the same note, slightly detuned against each other. The result is a thick, wide sound that many leads and basses need. The price: each unison voice uses a voice from the pool. A six-voice synth in full unison effectively plays mono. The Memorymoog could run up to 18 oscillators in unison in its mono mode, and in software synths you often set the number yourself.

Layer and split

A layer stacks two sounds, a split divides the keyboard into two zones with different sounds, such as bass on the left and lead on the right. Both only work if the synth can produce more than one sound at a time. Again: in a layer, every note uses two voices.

Multitimbrality and bi-timbrality

Multitimbral means a device produces several different sounds at once, usually addressed through separate MIDI channels. All multitimbral instruments are polyphonic, but not all polyphonic ones are multitimbral. This changed the home studio in particular: a single unit could play several instruments at once instead of needing several synths. Bi-timbral is the simplest form with two sounds. An example is the Sequential Prophet Rev2, which offers up to two sounds per program as a stack or split. More modern digital instruments divide into several parts or layers; the Korg Wavestate, for instance, plays four layers per performance.

Important for your math: the instrument's voices are shared among all sounds. A 16-voice bi-timbral synth running two sounds at once has fewer than 16 voices free for each. Whether each sound gets a fixed or a dynamic share differs by instrument.

Typical instruments as orientation (model data)

What does the difference cost?

Mono is cheap because only one signal path has to be built; on analog instruments each extra complete voice brings additional components. As orientation for new instruments (US street prices, rough values, they move often):

• Behringer Model D (monophonic): about $249 (as of Oct. 2026)
• Korg Minilogue xd (4 voices): about $680 to $800 depending on shop and promotion (as of Oct. 2026)
• Behringer DeepMind 12 (12 voices): about $800 to $860 at US shops (as of Oct. 2026); its list price at launch was $999.99
• Sequential Prophet Rev2 16-voice desktop: roughly $2,200 to $2,500 (as of Oct. 2026, varies by shop)

The pattern: four voices already cost a multiple of a mono unit, but the step from four to twelve voices is comparatively small in these examples. Vintage is different, since collector value rules. Used Roland Juno-106 units are typically listed in the US at about $1,500 to $2,000, with fully restored examples going well above that (as of Oct. 2026). For historical reference, the 1978 Prophet-5 launched at $3,995 and the 1982 Memorymoog at $4,795; the 2020 Prophet-5 reissue was announced at $3,499 for 5 voices and $4,299 for 10 voices. What to watch for with vintage polysynths is covered in buying a used synthesizer.

Buying decision: which type for which purpose?

• You want bass and leads: mono is perfectly enough, and you get more synthesizer for the money. Look at note priority and glide.
• You want chords, pads, piano-like playing: poly with at least four to eight voices; for long pads, 12 or more is better.
• You want pads with their own character on a tight budget: paraphonic can do, as long as you can live with a shared envelope.
• You want two sounds live: check split, layer and bi-timbrality, and count the voices per sound.
• You already own a mono synth: units with Polychain, or several instruments over MIDI, can be combined into a polyphonic setup, though it costs space, money and cabling. How the control works is explained in MIDI vs. CV/Gate.
• Digital or analog? Digital instruments scale voice count more easily, see analog, digital and virtual analog.

Check before you buy

1. Does the spec sheet say monophonic, paraphonic or polyphonic? With paraphonic, listen to how the envelope behaves.
2. How many voices in total, and how many are left in layer or unison?
3. When test-playing, hold a long pad chord and then keep playing quickly: do you hear notes being cut off?
4. On used polysynths, test every voice individually, because a dead voice often goes unnoticed in ensemble playing.

Myths and common mix-ups

• More oscillators mean more voices: wrong. The three oscillators in a Minimoog are one voice.
• Monophonic is inferior: mono is not a cut-down version but a role of its own. Many of the most famous basses and leads come from mono instruments.
• Voice stealing means the synth is broken: no, it is expected behavior at the voice limit.
• Paraphonic is the same as polyphonic: no, the shared filter and envelope clearly sound different.
• More voices are always better: only if you need them. For bass and lead they are pure reserve.
• Polyphonic means multitimbral: no. A six-voice Juno-106 plays one sound on all voices.
• The voice count on the spec sheet always applies: unison, layers and effects can lower the usable number.