Skip to main navigation Skip to search Skip to main content

Piano "forte pedal" analysis and detection

  • Antony Schutz
  • , Nancy Bertin
  • , Dirk T.M. Slock
  • , Bertrand David
  • , Roland Badeau

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Pressing the sustain pedal is one of the main musical gestures in a piano performance. It influences the sound produced by the instrument, and consequently the efficiency of any automatic system analyzing it. This paper aims at highlighting several features, observable on piano sounds played with pedal, which could be useful in the task of detecting the sustain pedal, and analyzing notes played while it is pressed. Here, we consider two hypotheses, derived from physical acoustics considerations and signal observations, that could help discriminate between notes played with and without the sustain pedal. First, the sustain pedal is found to increase the decay time of partials. This effects dominates the behavior of the partials, not only in duration but also in terms of spectral evolution. Secondly, when the sustain pedal is used, a noise floor appears, for all notes of the piano. Those two effects are investigated in this work, in particular through a method based on a "harmonic plus noise" decomposition. The possibility to use those features as a base for a detection system is discussed.

Original languageEnglish
Title of host publicationAudio Engineering Society - 124th Audio Engineering Society Convention 2008
Pages1064-1070
Number of pages7
Publication statusPublished - 1 Dec 2008
Event124th Audio Engineering Society Convention 2008 - Amsterdam, Netherlands
Duration: 17 May 200820 May 2008

Publication series

NameAudio Engineering Society - 124th Audio Engineering Society Convention 2008
Volume2

Conference

Conference124th Audio Engineering Society Convention 2008
Country/TerritoryNetherlands
CityAmsterdam
Period17/05/0820/05/08

Fingerprint

Dive into the research topics of 'Piano "forte pedal" analysis and detection'. Together they form a unique fingerprint.

Cite this