The Evaluation of Vocal Pitch Accuracy

Abstract

THE OBJECTIVE ANALYSIS OF WESTERN OPERATIC singing voices indicates that professional singers can be particularly ''out of tune.'' This study aims to better understand the evaluation of operatic voices, which have particularly complex acoustical signals. Twentytwo music experts were asked to evaluate the vocal pitch accuracy of 14 sung performances with a pairwise comparison paradigm, in a test and a retest. In addition to the objective measurement of pitch accuracy (pitch interval deviation), several performance parameters (average tempo, fundamental frequency of the starting note) and quality parameters (energy distribution, vibrato rate and extent) were observed and compared to the judges' perceptual rating. The results show high intra and interjudge reliability when rating the pitch accuracy of operatic singing voices. Surprisingly, all the parameters were significantly related to the ratings and explain 78.8% of the variability of the judges' rating. The pitch accuracy evaluation of operatic voices is thus not based exclusively on the precision of performed music intervals but on a complex combination of performance and quality parameters.

“I know it when I hear it”

“…As an example, the proportion of singers demonstrating poor singing abilities reaches more than 50% when the threshold is set at 50 cents (Loui, Demorest, Pfordresher, & Iyer, 2015; Pfordresher & Larrouy-Maestri, 2015). Therefore, one can assume that several factors linked to the performer (see Larrouy-Maestri, in press, for a review), his or her attire (Griffiths, 2010) or behavior (Thompson & Russo, 2007; Waddell & Williamon, 2017), the mode of presentation (i.e., visual and/or auditory, Tsay, 2013) or listeners’ expectations (Anglada-Tort & Müllensiefen, 2017; Kroger & Margulis, 2017) might modulate correctness category. The effects of non-musical variables on listeners’ tolerance would have to be specifically examined to adapt the definition of correctness to more natural settings.…”

“…The definition of correctness outlined here might not be adequate for highly trained voices, such as operatic voices. Indeed, such an acoustic signal is highly complex (Larrouy-Maestri, Magis, & Morsomme, 2014a) and the notion of pitch accuracy is not only based on pitch deviations between tones but relies on the association of several parameters such as energy distribution and vibrato (Larrouy-Maestri, Magis, & Morsomme, 2014b; Larrouy-Maestri, Morsomme, Magis, & Poeppel, 2017). In addition, specific tones might be purposely mistuned for expressive purpose (Sundberg et al, 2013).…”

Perceptual (but not acoustic) features predict singing voice preferences

“…Acoustic analyses. Acoustic analysis followed the procedure extensively described in (18,50,51), using AudioSculpt 2.9.4v3 software, OpenMusic 6.3 software (IRCAM, Paris, France) and Praat (52) (Version 6.0.46). The following parameters were analysed: pitch interval deviation (average of pitch deviations between the expected and performed intervals along the melody); vibrato rate (number of quasiperiodic modulations of the F0 per second, in Hz); vibrato extent (amplitude of the F0 variations within the same tone, in cents); energy distribution (energy of the 2.4-5.4 kHz band divided by the energy across the 0-10 kHz band -the rationale behind this computation is that a high score in the energy distribution variable indicates a strong reinforcement of the band containing the singer's formant); jitter (local; perturbation in the F0 from cycle to cycle); shimmer (local; perturbation in the amplitude of the F0 from cycle to cycle); average tempo (excerpt's length divided by the number of beats).…”

“…Particularly for the singing voice, there is evidence that specific acoustic features influence the perception of pitch accuracy, or of singing performances as "correct" (18) and that the use of such features is affected by the music expertise of the listeners (19,20) -but such research does not answer questions about singing preferences. In fact, aesthetic appreciation probably goes beyond correctness, that is to say, people do not attend concerts to hear 'correct' performances, but to enjoy them.…”

An alternative m-learning scoring approach to automatically evaluate Iqro’s pronunciation based on pitch, volume, and rhythm features

“…With that condition, the improvement level is difficult to trace. Some studies have been working in the scientific field of performance analysis of the voice with acoustic features as parameters [7]- [10]. Previous research has proposed an automatic singing evaluating system based on acoustic features and rhythm [11].…”