Objective:
Previous research across acoustics, voice medicine, and vocal performance has investigated differences among Chinese National Singing, Bel Canto, and Popular Music, primarily focusing on spectral features, resonance patterns, and the singer’s formant. However, these studies often relied on single-dimensional acoustic parameters, leaving the multidimensional nature of vocal production insufficiently explored.
This study introduces a multi-metric voice mapping framework based on the Voice Range Profile (VRP), integrating acoustic parameters, electroglottographic (EGG) measures, and visualized mapping techniques. Through this approach, we aim to quantitatively characterize and compare the vocal mechanisms underlying the three singing styles, reveal their acoustic interrelations, and provide empirical foundations for both singing voice theory and vocal pedagogy.

Method:
This study adopts a cycle-by-cycle analysis, where each phonatory cycle is quantified through multiple voice quality metrics conditioned on fundamental frequency (F₀) and sound pressure level (SPL).
Acoustic and EGG data were collected in the Laboratory of Language Sciences, Peking University (Beijing, China) and then analyzed using the FonaDyn software and customized python scripts. A set of metrics is encompassed: pitch in MIDI and loudness in SPL, voice quality metrics (clarity, crest factor, spectral balance, CPP, sample rate entropy), and vocal fold contact metrics obtained by EGG (contact quotient, maximum derivative of EGG signal). Statistical analyses included both parametric (ANOVA) and non-parametric (Kruskal–Wallis) tests, depending on data distribution, followed by effect size estimation and contribution scoring to evaluate the relative discriminative power of each metric across singing styles and genders.

Conclusion:
Key findings reveal clear acoustic distinctions across both style and gender. Sample entropy stands as the most influential differentiating parameter (contribution score = 7.96), with Bel Canto singing exhibiting the most regular vocal patterns and Popular Music the most complex. Crest factor and contact quotient are secondary but significant contributors, reflecting differences in vocal energy and glottal closure strength. Overall, all selected metrics show significant differences among styles, forming a consistent acoustic gradient: Bel Canto voices are the most stable and resonant, followed by Folk Songs in an intermediate position, and Popular Music displays the highest variability and vocal complexity.