Putting the Singing Voice on the Map : Towards Improving the Quantitative Evaluation of Voice Status in Professional Female Singers

Abstract: Diagnostic and evaluative methods used in voice care are mostly designedfor the speaking voice, and are not necessarily directly applicable to thesinging voice. This thesis investigated the possibilities of fine tuning, improvingand quantifying the voice status assessment of the singer, focusingespecially on the Western operatic female voice. In Paper I, possible singer-specific Voice Range Profile (VRP) characteristicsand tasks were explored and VRP data for 30 professional female Western opera singers was collected. Vocal productions were controlled for a physiological VRP (VRPphys) and for a stage performance context (VRPperf) and outcome differences were identified. Task design was critical for the(VRPphys) but had very little effect on the VRPperf. Significant voice category differences (between soprano,mezzo-soprano and contralto) were limited to frequencyrelated metrics. Two new VRP metrics, the area above 90 dB (Perc90dB) and the sound pressure level extent (SPLext), were found to be key metrics to the study of VRPs for singers. Paper II investigated, in conjunction with the VRP, whether the sound pressure level (SPL) or the skin acceleration level (SAL) was more correlated to the subglottal pressure (Ps). SAL was much less F0 dependent than SPL and facilitated the interpretation of VRP data. However, the correlation between SAL and Ps was found to be weaker than that between SPL and Ps. Papers III and IV explored the mapping of self-perceived impairmentrelated difficulties into the VRP. A modified phonetograph was tested first with a healthy singer population and then with a singer-patient group. Subjects used a button device to communicate their self-perceptions while singing, and were consistent in task replications as well as across different tasks. Healthy singers pressed mostly at the extreme limits of the VRP, where loss of vocal control could be expected and their presses were mostly concentrated on the periphery of the VRP area. Singer patient button- press patterns were distinct from patterns observed in healthy singers. Singer patients pressed mainly inside the VRP boundaries, in the higher range and at intermediate intensities. In Paper V, the Voice Handicap Index for singers was translated and adapted to Swedish (Röst Handikap Index för sångare or RHI-s). The questionnaire was found to be a reliable and a valid instrument. High correlations between general perceptual patient VAS ratings and the questionnaire scores underscored the instrument’s internal coherence. Overall, patient scores (including subscales) were significantly higher than healthy singer scores. The results showed implicitly the necessity and usefulness of adapting clinical procedures to specific patient populations.Together, the results of these five papers can ultimately be of value tovoice clinicians who are treating singers. The results obtained also contributeto the understanding of the singing voice and underline the importance ofproperly documenting the singing voice.