Audio-Perceptual Features of the Singing Voice Classified by 4 Judges’ Groups: Development of an Appreciation Scale

Mendes, Ana; Vaz, Inês; Ibrahim, Soraia

http://hdl.handle.net/10400.26/10625

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
2015 Audio-perceptual features of the singing voice.pdf		384.15 KB	Adobe PDF	Download

Send Feedback

Authors

Mendes, Ana

Vaz, Inês

Ibrahim, Soraia

Abstract(s)

Audio-perceptual voice analysis is common in clinical practice. Eventhough acoustic analysis provides objective and useful data, audioperception is irreplaceable. The main purpose of this observational research was to develop a singing voice rating scale (Escala de Aprecia¸c˜ao da Voz Cantada – EAVOZC) and second to identify the Fado’s singing voice audio-perceptual profile. EAVOZC has a continuous visual analogue scale with a 50mm line to grade the accentuation level on 13 audio-perceptual voice parameters: 1) pitch, 2) loudness, 3) resonance, 4) phonatory and respiratory coordination, 5) precise articulation, 6) voice quality (roughness, breathiness, strain and asthenia), 7) vibrato, 8) voice projection, 9) voice emotional expression, 10) timbre, 11) brilliance, 12) tuning and 13) global voice appreciation. This paper reports the audio-perceptual classifications of singing voice using the EAVOZC. 40 judges participated in this study: 10 singing teachers (ST), 10 speech-language pathologists (SLP), 10 Fado singers (FS) and 10 naive listeners (NL). Age ranged from [19–73] yrs. Judges listened to 20 voice samples of the Fado Chorus “Nem ´as paredes confesso”. These samples were produced by 10 males, 10 females; 10 amateur, 10 professional singers divided in two age groups [18–55] and [55–70]. Judges filled in the EAVOZC. One-way ANOVA and t-test were used with a a at .05. Results revealed ST and SLP significantly judged singing voice samples in a similar manner in terms of the audio-parameters above listed. The same happened in the groups FS and NL except for timbre and brilliance. Therefore, from now on the results were presented based on 2 groups instead of 4. EAVOZC parameters showed thatmale and female voices’ were distinct in roughness, pitch, tuning and global voice appreciation. Professional Fado singers had higher scores in tuning, emotional expression, precise articulation and had less tension when compared to amateurs. Timbre was classified as dark on male voices and light on female voices. Brilliance was absent in males and present in female voices.