A measurement of piano playing
Whether higher notes are played louder.
The same passage — Schubert Op. 90/3, bars 1–2 — 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.
Slope of a regression of melody velocity on pitch, in velocity units per octave.
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.
| Piece | Recordings | Lowest | Median | Highest |
|---|---|---|---|---|
| Bach Prelude in C | 27 | -4.8 | 0.5 | 4.9 |
| Chopin Étude 10/1 | 31 | 0.0 | 1.4 | 3.7 |
| Nocturne Op. 9/2 | 32 | -6.0 | 4.0 | 9.9 |
| Ravel Pavane | 33 | -0.1 | 4.1 | 10.1 |
| Chopin Ballade 4 | 12 | 4.0 | 5.0 | 7.3 |
| Chopin Étude 10/12 | 19 | 2.9 | 5.1 | 7.1 |
| Ravel Ondine | 61 | 1.8 | 5.7 | 8.7 |
| Schubert Op. 142/3 | 9 | 3.9 | 6.0 | 7.7 |
| Mozart K. 545 | 37 | 2.7 | 7.2 | 12.4 |
| Rachmaninoff Prelude | 29 | 4.6 | 7.6 | 10.3 |
| Fantaisie-Impromptu | 20 | 6.3 | 9.1 | 11.8 |
| Brahms Op. 118/2 | 28 | 2.8 | 9.6 | 13.8 |
| Clair de lune | 27 | 6.8 | 9.7 | 12.1 |
| Schubert Op. 90/3 | 43 | 2.8 | 10.1 | 16.2 |
| Gymnopédie 1 | 23 | 6.9 | 10.7 | 13.8 |
| Träumerei | 31 | 4.0 | 10.7 | 16.5 |
| Pathétique II | 47 | 8.4 | 12.2 | 16.0 |
| Nocturne Op. 48/1 | 27 | 8.0 | 12.3 | 15.1 |
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 1.62. 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.