Abstract
The majority of quantitative research on the singing voice has been undertaken in a laboratory environment, where analysis tools are frequently based on phonetically balanced texts and vowel vocalisations rather than ‘real’ repertoire. Most singers of music in the Western Classical tradition rely on the acoustics of the performance venue to provide aural feedback to control aspects of their vocal performance, and are therefore particularly vulnerable to changes in acoustic environment.
This pilot project examined quantitative measures of the performance of a singer in two different environments. The project analyzed a sung phrase of ~50 seconds duration taken from recordings by a solo unaccompanied mezzo-soprano singer performing in a hemi-anechoic chamber (RT60 = 0), and the same singer recorded in a large Victorian church with RT60 ~ 3.5 s. The phrase was recorded several times in each environment, and was part of a solo well known by the singer. A number of key parameters were analysed, including Mean frequency over a number of sustained notes, mean tempo, vibrato rate/depth, and pitch of formant frequencies of sustained vowels in the musical phrase evaluated using Long Term Average Spectra.
Significant variation was found between the recordings in each venue for the parameters of tempo, vibrato rate and depth, and the frequencies of formants F1-F3. Pitch variation was exhibited but not at a significant level.
Howard and Brereton (2008) analysed singers in a similar manner, examining different parameters, and found raised vocal fold closed quotients, a shallower spectral slope and increased intensity for the anechoic recordings, suggesting heightened effort and increased vocal load and a resultant change in singer tone quality.
This study supports these findings, indicating that significant changes in tone quality, and other key aspects of performance are highly likely where singers are assessed in anechoic spaces, and that researchers need to take into account this consideration when assessing vocal properties.
This pilot project examined quantitative measures of the performance of a singer in two different environments. The project analyzed a sung phrase of ~50 seconds duration taken from recordings by a solo unaccompanied mezzo-soprano singer performing in a hemi-anechoic chamber (RT60 = 0), and the same singer recorded in a large Victorian church with RT60 ~ 3.5 s. The phrase was recorded several times in each environment, and was part of a solo well known by the singer. A number of key parameters were analysed, including Mean frequency over a number of sustained notes, mean tempo, vibrato rate/depth, and pitch of formant frequencies of sustained vowels in the musical phrase evaluated using Long Term Average Spectra.
Significant variation was found between the recordings in each venue for the parameters of tempo, vibrato rate and depth, and the frequencies of formants F1-F3. Pitch variation was exhibited but not at a significant level.
Howard and Brereton (2008) analysed singers in a similar manner, examining different parameters, and found raised vocal fold closed quotients, a shallower spectral slope and increased intensity for the anechoic recordings, suggesting heightened effort and increased vocal load and a resultant change in singer tone quality.
This study supports these findings, indicating that significant changes in tone quality, and other key aspects of performance are highly likely where singers are assessed in anechoic spaces, and that researchers need to take into account this consideration when assessing vocal properties.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of Physiology and Acoustics of Singing conference 2010 |
| Editors | David Howard, Jude Brereton |
| Publisher | University of York |
| Publication status | Published - 2010 |