Why Does Digital Piano Polyphony Matter?

An acoustic piano has 88 keys – that’s the available polyphony. So why does a digital piano need 120 voices and more – when you only have ten fingers? How does all that work? And most importantly, how much digital piano polyphony do you need to play the piano well?

How do “piano polyphony” and “digital piano polyphony” differ?

The most common sampling technique does not simulate the vibrations of a piano string – this would be the physical modeling technique. Sampling means that only a digitalized recording of a single acoustic event is reproduced. Simulating the spectacular dynamic response of a grand piano therefore places high demands on piano sampling.

The piano sound consists of many individual samples to create a dynamically playable instrument. This is the big difference to an acoustic piano: the string can be struck and then vibrate over and over again. Alone the repetition of a single note when playing with the sustain pedal can require a high level of digital piano polyphony. This further increases with the complexity of sound generation.

Samples per voice…

The number of voices available to play on the digital piano depends on several factors. For example, the type of sound generation, the playing mode (Dual, Layer), and the complexity of sound details. One factor that often causes confusion: To play a stereo-sampled piano sound, you basically need two (!) voices for each note you play, since a stereo sample consists of two samples – one for the left and one for the right channel.

Why is digital piano polyphony a dynamic value?

Once again, stereo sampling means combining two voices into one playable sound. But what about all the details that make the piano sound so authentic? The lift and touch of the dampers, the strike of the hammers… These, of course, require samples you can hear at the same time as the piano sound. Ergo, the more complex the sound generation, the greater the need for voices. Of course, all this is at the expense of the total number of voices available for playing.

Then there is the DUAL or LAYER mode, which layers two sounds on top of each other. If we use two stereo sounds for layering – say, piano plus strings – we need a total of four voices to produce the overall sound. A four-voice chord now “consumes” 16 voices. For 120-voice polyphony, this means: 120 divided by 4 – you still have 30 voices available to play the piano thanks to layer bundling.

Digital piano polyphony: Dynamic voice distribution

Fortunately, digital pianos work with a sophisticated logic that distributes the available voices evenly. Certain notes are prioritized right away – there are different criteria to implement this. If the polyphony is exhausted, other, ‘non-priority’ notes end so new notes can be played. For example, bass notes always have priority: they are the foundation of the sound and should continue to sound as long as possible. In the higher registers, notes fade away faster anyway, so polyphony voices become available again more quickly. This happens even if you are still using the sustain pedal.

This dynamic distribution of the voices ensures that the overall sound is (almost) always well-balanced, even if you play repetitions or long arpeggios using the sustain pedal. In general, however, the following applies: The lower the digital piano polyphony, the more limits are there to the dynamic distribution of voices.

What to know about sound complexity

The authentic playing and sound characteristics of a digital piano place high demands on electronic sound generation: the more authentic, the more complex the sound generation. The big difference between an acoustic piano and an electronic piano becomes clear when you hear the repetition of a single note while using the sustain pedal.

In an acoustic piano, the hammers only strike the strings of the note again. The electronic reproduction in the digital piano requires two notes: When the second note is struck, the first should not be cut off, but should fade into the background.

Now you can imagine how long repetitions can affect voice consumption when playing with the sustain pedal. If you consider the normal interaction with a stereo-sampled sound, a polyphony of 64 voices, for example, which seems sufficient at first, is quickly put into perspective. There are 32 playable voices available. If you sustain only four notes via the sustain pedal, you need eight voices for the decay. Now play two 6-note piano voicings using the pedal, and the dynamic voice distribution must come into action.