Criteria and tools for objectively analysing the vocal accuracy of a popular song

in Logopedics, Phoniatrics, Vocology (in press)

Abstract This study aims to validate our method for measuring accuracy in a melodic context. We analysed the popular song “Happy Birthday” sung by 63 occasional and 14 professional singers thanks to ... [more ▼]

Abstract This study aims to validate our method for measuring accuracy in a melodic context. We analysed the popular song “Happy Birthday” sung by 63 occasional and 14 professional singers thanks to AudioSculpt and OpenMusic (Ircam, Paris, France). In terms of evaluation of the pitch interval deviation, we replicated the profile of occasional singers described in the literature (the slower the performance, the more accurate it is). Our results also confirm that the professional singers sing more accurately than occasional singers but not when a Western operatic singing technique is involved. These results support the relevance of our method for analysing vocal accuracy of occasional and professional singers and led us to discuss adaptations to be implemented for analysing the accuracy of operatic voices. [less ▲]

The singing voice evaluation

Conference (2013, March 01)

Singing is an activity widely shared. In order to determine whether a vocal performance is “in tune” or not, to understand better the causes of poor pitch singing or to judge the quality of a singer ... [more ▼]

Singing is an activity widely shared. In order to determine whether a vocal performance is “in tune” or not, to understand better the causes of poor pitch singing or to judge the quality of a singer, objective tools are needed. Studies leaded for a PhD validate an objective analysis of the vocal accuracy but also observe the effects to take into account for the singing voice evaluation. Experimental as well as artistic and pedagogical perspectives appear from the conclusions of those studies. [less ▲]

The evaluation of singing voice accuracy: A comparison between subjective and objective methods

in Journal of Voice (2013), 27(2), 2591-5

Objective: Vocal accuracy of a sung performance can be evaluated by two methods: acoustic analyses and subjective judgements. Acoustic analyses have been presented as a more reliable solution but both ... [more ▼]

Objective: Vocal accuracy of a sung performance can be evaluated by two methods: acoustic analyses and subjective judgements. Acoustic analyses have been presented as a more reliable solution but both methods are still used for the evaluation of singing voice accuracy. This paper presents a first time direct comparison of these methods. Methods: 166 untrained singers were asked to sing the popular song « Happy Birthday ». These recordings constituted the database analyzed. Acoustic analyses were performed to quantify the pitch interval deviation, the number of contour errors and the number of tonality modulations for each recording. Additionally, eighteen experts in singing voice or music rated the global pitch accuracy of these performances. Results: A high correlation occurred between acoustic measurements and subjective rating. The total model of acoustic analyses explained 81% of the variance of the judges’ scores. Their rating was influenced by both tonality modulation and interval deviation. Conclusions: This study highlights the congruence between objective and subjective measurements of vocal accuracy within this first time comparison. Our results confirm the relevance of the pitch interval deviation criterion in vocal accuracy assessment. Furthermore, the number of tonality modulations is also a salient criterion in perceptive rating and should be taken into account in studies using acoustic analyses. [less ▲]

Trained Singers Corpus

Computer development (2012)

Two songs (Happy Birthday and a chosen romantic melody) were performed by 50 trained singers aged from 19 to 66 years old (on average: 36.94). These 38 women and 12 men began their musical education ... [more ▼]

Two songs (Happy Birthday and a chosen romantic melody) were performed by 50 trained singers aged from 19 to 66 years old (on average: 36.94). These 38 women and 12 men began their musical education between 6 and 49 years old (on average: 20.18) and have a musical background of 5 to 51 years (on average: 19.86). They practice 13.68 hours per week, on average. After the production of two glissandi, the participants were asked to sing the French version of the popular tune “Happy Birthday”, without a compulsory tonality, two times. For the first time they were instructed to sing "naturally" (sequence a) and then to sing with an operatic singing technique "like on stage" (sequence b). Each participant chose a romantic melody from their musical repertory and sang this melody three times: "naturally" (sequence c), with an operatic singing technique (sequence d) and with closed mouth "humming" (sequence e). The performance order was similar for each trained singer (sequence a, e, c, d, b). [less ▲]

Impact du stress sur la justesse d’une performance vocale chez des étudiants de conservatoire

Conference (2012, October 15)

Un haut niveau d’anxiété entraîne des symptômes physiques et cognitifs. Cet état émotionnel peut se répercuter sur les caractéristiques acoustiques de la voix. Cette étude vise à observer l’impact d’une ... [more ▼]

Un haut niveau d’anxiété entraîne des symptômes physiques et cognitifs. Cet état émotionnel peut se répercuter sur les caractéristiques acoustiques de la voix. Cette étude vise à observer l’impact d’une situation stressante sur la justesse d’une performance vocale. 31 étudiants (deux niveaux différents) recrutés au sein des Conservatoires Royaux de Belgique ont chanté une mélodie à trois moments de l’année : deux fois en situation calme (T1 et T3) et une fois en situation anxiogène (T2), c’est à dire le jour de l’examen. Nous avons contrôlé le degré d’anxiété par la mesure de la fréquence cardiaque. Un questionnaire nous renseigne sur l’intensité des symptômes cognitifs et somatiques perçus par les étudiants et leur direction (altérant ou facilitant). La précision des intervalles et le maintien de la tonalité ont été mesurés pour chaque production. Nous comparons les scores obtenus lors d’une situation calme et lors de l’examen. Nous observons que la fréquence cardiaque augmente significativement en situation anxiogène (F(2,90) = 12.60 ; p < .001). L’intensité des symptômes cognitifs et somatiques est plus élevée lors de la situation anxiogène mais l’évolution des symptômes somatiques entre T3 et T2 est plus importante pour les étudiants de 2ème année par rapport aux étudiants de 1ère année. Concernant la justesse, les résultats indiquent qu’un état de stress induit une meilleure précision des intervalles chez les étudiants de 1ère année (p = .02) mais altère le maintien de la tonalité chez les élèves de 2ème année (p = .01). En conclusion, une situation d’examen entraîne une augmentation de la fréquence cardiaque mais l’interprétation des symptômes somatiques dépendrait du niveau d’étude. Les conséquences d’une situation anxiogène, tel un examen de solfège, sur la justesse d’une performance vocale, peuvent être bénéfiques mais sont négatives lorsque l’enjeu est important. [less ▲]

A comparison between subjective and objective methods for evaluating the vocal accuracy of a popular song

Poster (2012, July 25)

Vocal accuracy of a sung performance can be evaluated by two methods: acoustical analyses and subjective judgements. For one decade, acoustic analyses have been presented as a more reliable solution to ... [more ▼]

Vocal accuracy of a sung performance can be evaluated by two methods: acoustical analyses and subjective judgements. For one decade, acoustic analyses have been presented as a more reliable solution to evaluate vocal accuracy, avoiding the limitation of experts’ perceptive system and their variability. This paper presents for the first time a direct comparison of these methods. 166 occasional singers were asked to sing the popular song « Happy Birthday ». Acoustic analyses were performed to quantify the pitch interval deviation, the number of contour errors and the number of tonality modulations for each recording. Additionally, eighteen experts in singing voice or music rated the global pitch accuracy of these performances. The results showed a high inter-rater concordance within the judges. In addition, a high correlation occurred between acoustic measurements and subjective rating. The total model of acoustic analyses explained 81% of the variance of the judges’ scores. Their rating was influenced by both tonality modulations and interval deviations. This study highlights the congruence between objective and subjective measurements of vocal accuracy while the assessment is done by music or singing voice experts. Our results confirm the relevance of the “pitch interval deviation” criterion in vocal accuracy assessment. Furthermore, the “number of tonality modulations” is a salient criterion in perceptive rating and should be taken into account in studies using acoustical analyses. [less ▲]

The relationship between vocal abilities and singing accuracy

Conference (2012, May 10)

Poor-pitch singing could be caused by poor pitch perception or poor vocal-motor control abilities. This study aims to contrast these two possible causes in order to determine the role of vocal control on ... [more ▼]

Poor-pitch singing could be caused by poor pitch perception or poor vocal-motor control abilities. This study aims to contrast these two possible causes in order to determine the role of vocal control on the accuracy of sung performances among nonmusicians. Participants matched recordings of their own voices either by singing, or by manipulating those recordings on a physical instrument which could control the pitch of the vocal recording playback by sliding the finger along a position sensor. In addition, participants sang a full song from memory. Overall, participants were more accurate at matching the pitch of the original recording with the instrument than with their voice. In addition, singers who were more accurate at vocal pitch matching tended to have better vocal quality (as assessed through standard measurements, e.g. jitter, stability), and were better at singing whole songs. This pattern of results confirms that vocal-motor control, rather than pitch perception ability, is the primary driver of singing ability, and provides insight into the relationship between pitch accuracy and vocal quality. [less ▲]

Critères et outils d'analyse de la justesse vocale

Scientific conference (2011, November 17)

La justesse est un indicateur de la qualité d’un chant et se base principalement sur la précision de l’intonation. Cette étude vise à valider des critères et outils de mesure de la justesse par l’analyse ... [more ▼]

La justesse est un indicateur de la qualité d’un chant et se base principalement sur la précision de l’intonation. Cette étude vise à valider des critères et outils de mesure de la justesse par l’analyse d’une chanson populaire produite par des chanteurs occasionnels et des chanteurs professionnels. Nous avons analysé les productions vocales de 63 chanteuses occasionnelles et de 14 chanteuses ayant suivi une formation au chant lyrique. Les sujets ont chanté « Joyeux anniversaire ». Les chanteuses professionnelles l’ont chanté une première fois sans technique vocale particulière (ST) et une seconde fois avec une technique vocale lyrique (AT). La partie stable de chaque voyelle a été moyennée afin d’estimer la hauteur fréquentielle de chaque note. Les erreurs de type « précision des intervalles » et « maintien de la tonalité » ont pu être observées. Les résultats montrent une corrélation entre le tempo de la production et la précision des intervalles (r=.32 ; p=.01), ce qui réplique le profil des chanteuses occasionnelles décrit dans la littérature. Ils confirment également que les chanteuses professionnelles chantent plus juste que les chanteuses occasionnelles pour la précision des intervalles (t=4,51 ; p<.001) et le maintien de la tonalité (t=3,05 ; p=.004). Ces résultats contribuent à valider la pertinence de nos critères et outils d’analyse de la justesse vocale. Par ailleurs, nous observons que les chanteuses AT sont particulièrement fausses selon le critère de « précision des intervalles », ce qui nous conduit à discuter des adaptations à appliquer pour l’analyse de la justesse de ces voix. [less ▲]

Happy Birthday Corpus

Computer development (2011)

Popular song sung by 166 french occasionnal singers. The participants sang the French version of the popular tune “Happy Birthday”, without a compulsory tonality, after production of two glissendi (a ... [more ▼]

Popular song sung by 166 french occasionnal singers. The participants sang the French version of the popular tune “Happy Birthday”, without a compulsory tonality, after production of two glissendi (a continuous glide from a low note to a high note and vice versa). The aim of these glissendi was to warm up the vocal organs, to verify the vocal capacity of the subjects and to encourage a lack of inhibition in front of the experimenter and the recording equipment. The sound recordings were made using a microphone (Sennheiser HS2) and a digital recorder (Marantz Professional Solid State Recorder PMD67). [less ▲]

Vocal accuracy in the singing voice

Scientific conference (2011, May 27)

L'état de l'art dans le domaine de la "justesse vocale" et de ses troubles permet de mieux comprendre la pertinence de la recherche scientifique dans ce domaine. Les premiers résultats confirment l ... [more ▼]

L'état de l'art dans le domaine de la "justesse vocale" et de ses troubles permet de mieux comprendre la pertinence de la recherche scientifique dans ce domaine. Les premiers résultats confirment l'utilité d'une évaluation objective de performances vocales. Une comparaison entre les méthodes utilisées au sein de différents laboratoires permet d'améliorer la qualité scientifique de l'évaluation vocale et ainsi d'envisager les implications cliniques de ces recherches. [less ▲]