Title :
Inter-dependence of spectral measures for the singing voice
Author :
Mesaros, Annamaria ; Astola, Jaakko
Author_Institution :
Dept. of Electron. & Telecommun., Cluj-Napoca Tech. Univ., Romania
Abstract :
This paper deals with investigating the problem of singer identification, by studying some spectral features and the way they depend on each other and on musical performance. Features like the spectral centroid, the tristimulus, the odd/even harmonics relationship are calculated and their evolution during a musical phrase is studied. Their inter-dependence is tested using the Spearman´s correlation coefficient. The mentioned features have previously found to be useful for musical instruments description. The results we obtained suggest that the spectral characteristics of the singing voice are mostly determined by the position of the formants, under conscious control of the singer.
Keywords :
audio signal processing; musical instruments; speaker recognition; spectral analysis; Spearman correlation coefficient; musical instrument; musical performance; musical phrase; odd/even harmonics relationship; singer identification; singing voice; spectral centroid; spectral characteristic; spectral measure inter-dependence; Humans; Instruments; Medical diagnostic imaging; Multiple signal classification; Pathology; Signal analysis; Signal processing; Speech analysis; Speech recognition; Timbre;
Conference_Titel :
Signals, Circuits and Systems, 2005. ISSCS 2005. International Symposium on
Print_ISBN :
0-7803-9029-6
DOI :
10.1109/ISSCS.2005.1509915