A measurement of piano playing

Contour smoothness

How smoothly the melody's loudness changes from note to note.

What it sounds like

The same passage — Bach Prelude in C, bars 1–3 — from the recordings that sit lowest and highest on this measurement. Every clip is a note list replayed on one sampled grand at one gain, so the only difference you are hearing is the playing.

Least

Sviatoslav Richter, 1970 — bars 1–3, contour smoothness 0.47 0–110 s
Alice Sara Ott, 2021 — bars 1–3, contour smoothness 0.48 0–18 s

Most

Angela Hewitt, 2008 — bars 1–3, contour smoothness 0.86 0–111 s
Till Fellner, 2004 — bars 1–3, contour smoothness 0.84 0–110 s

How it is calculated

Share of melody-velocity variance surviving a five-note moving average.

How much pianists differ on it

Every piece in the library, with the range across its recordings. The units mean the same thing in every row, except for basic tempo, which is in each piece’s own beat.

PieceRecordingsLowestMedianHighest
Chopin Étude 10/1310.250.460.70
Ravel Pavane330.490.560.67
Träumerei310.410.570.71
Schubert Op. 90/3430.420.610.76
Chopin Étude 10/12190.440.620.73
Gymnopédie 1230.450.630.75
Schubert Op. 142/390.630.660.72
Bach Prelude in C270.470.670.86
Ravel Ondine610.550.690.82
Mozart K. 545370.520.690.77
Brahms Op. 118/2280.560.700.78
Pathétique II470.640.710.78
Nocturne Op. 9/2320.560.740.84
Chopin Ballade 4120.710.740.79
Nocturne Op. 48/1270.600.750.81
Clair de lune270.670.750.85
Fantaisie-Impromptu200.690.800.88
Rachmaninoff Prelude290.740.860.93

How reliably it can be measured

A Disklavier capture — the pianist’s own key sensors, no microphone — was rendered to audio, transcribed as if it were a recording, and re-measured. The error the transcription step introduces, as a multiple of how far pianists actually differ on this dial, is 2.47. noisy The transcription moves this dial about as far as pianists differ on it, so no claim should rest on where a single recording sits.