A Voice Mapping Based Comparative Study of Vocal Characteristics in Bel Canto, Chinese National Singing and Popular...

This study aims to investigate stylistic differences in singing styles by constructing two-dimensional voice maps. Rather than relying solely on global mean comparisons, the analysis examines how multiple complementary voice quality metrics are distributed within a structured phonatory space defined by fundamental frequency and sound pressure level.

Acoustic and electroglottographic (EGG) signals were recorded simultaneously from singers performing Bel Canto, Chinese National Singing, and Popular Singing. Core metrics from acoustic and EGG signals were analyzed across f₀-SPL regions. Per-cycle statistics and one-way ANOVAs were conducted within each sex, and spatial distributions were visualized using two-dimensional voice maps.

At the level of global cycle-pooled means, one-way ANOVAs revealed significant style effects on CPP, spectrum balance (both P < 0.001), and Q_Δ (P < 0.05) in both sexes (Bel Canto highest, Popular Singing lowest, Chinese National Singing intermediate). Voice maps further showed that these differences are not uniformly distributed: Bel Canto's high values concentrate in mid-to-high f₀ × SPL, while Popular Singing extends into lower-SPL regions.

The findings demonstrate that stylistic differences are systematically embedded within the phonatory space and vary across vocal regions. Mean-based analyses provide a reference; voice maps provide the spatial context that gives those means their interpretive meaning.