Multidimensional timbre analysis of shakuhachi honkyoku - Aix-Marseille Université Accéder directement au contenu
Communication Dans Un Congrès Année : 2006

Multidimensional timbre analysis of shakuhachi honkyoku

Résumé

A method of analysing timbre in melody that takes account of its multidimensional characteristic is described. The aim of this analysis technique is to uncover the role of timbre as a structural element in melody. Gaining an understanding of how an abstract, multidimensional sound phenomenon such as timbre is used in a structured manner can lead to important insights into potential cognitive abilities; our ability to form mental representations of abstract phenomena so that they can be conserved and, subsequently, used in an organised or structured manner. This method of analysis comprises two main components, a first which carries out a time-frequency analysis of the melodic signal and processes the signal data using DSP techniques that model aspects of auditory processing before calculating the timbre descriptors, and a second that represents the melody in terms of its timbral changes rather than its absolute timbral values. A self-organising feature map is used to reduce the timbral detail and the dimensionality of the timbral representation, and to contrast-enhance the timbral changes. An example of the implementation of this analysis technique is presented using an extract from a Japanese shakuhachi honkyoku melody, chosen because of its accepted exploitation of timbre.
Fichier principal
Vignette du fichier
Multidimensional_timbre_analysis_of_shakuhachi_hon.pdf (305.12 Ko) Télécharger le fichier

Dates et versions

hal-03991696 , version 1 (16-02-2023)

Identifiants

  • HAL Id : hal-03991696 , version 1

Citer

Deirdre Bolger, Niall Griffith. Multidimensional timbre analysis of shakuhachi honkyoku. Conference on Interdisciplinary Musicology (CIM05), Society for Interdisciplinary Musicology, Mar 2005, Montreal (Canada), Canada. ⟨hal-03991696⟩
15 Consultations
19 Téléchargements

Partager

Gmail Facebook X LinkedIn More