Introduction. Registers are considered a fundamental yet controversial topic in relation to the lack of generally accepted definitions and lack of consensus on pedagogical application. Recent discoveries suggest that singers are able to use specific vocal tract configurations across upward of three octaves without register changes. Previous research has shown differences between registers at the glottal level as well as in perceptual, acoustic, physiological, and aerodynamic dimensions, while other studies have suggested that singers can maintain specific vocal configurations past previously observed register shifts. This study investigates several loudness and vocal quality conditions across 3-4 octaves.
Methods. Double-case study of two professional singers performing vocal tasks based on Complete Vocal Technique modes: sustained vowels at varying intensities (MP, F, and FF) across their vocal range (three octaves and four octaves, respectively). A total of 572 (F1) and 516 (M1) samples were collected and analyzed using MATLAB and Praat, focusing on sound pressure level (SPL), fo, formants (F1-F4), spectral parameters (SPR, L1-L2, and SCO-2.5), jitter, shimmer, smoothed cepstral peak prominence (CPPS), noise-to-harmonic ratio (NHR), and electroglottographic (EGG) analysis. Statistical methods included normality testing, t tests, Wilcoxon tests, Bonferroni correction, linear regression, and factor analysis.
Results. Both singers produced samples in M1- and M2-type vibratory mechanisms extending beyond typically observed pitch regions for traditional registers, with the male covering a total of four octaves (C2-C6) with all conditions represented in three octaves, while the female singer covered a total of three octaves (D3-D6) with all conditions represented in almost two octaves. Metal and density parameters systematically and linearly varied with fo and influenced SPL, acoustic spectra, and EGG. Observed vibrational patterns for fuller and reduced metallic and fuller and reduced density conditions resembled patterns previously reported for modal or M1 type of vibratory patterns, whereas falsetto conditions resembled patterns typically reported for M2 or head voice-type vibratory patterns. Fuller metallic conditions exhibited higher SPL and Qx, while non-metallic conditions showed lower SPL and EGG-waveform magnitude. Factorial analysis revealed distinct statistical differentiation for metal and density variations. Classical samples showed restricted use of vocal modes throughout the range, whereas CCM samples did not. The findings highlight the complex and nuanced interplay between glottal behavior and vocal tract setup to offer new perspectives on how singers navigate throughout their vocal range.
Conclusion. Metal and density parameters offer a stable, measurable, and statistically significant nuancing beyond traditional register classifications. Participants produced stable vocal configurations across more than three octaves. Register boundaries are fluid and singers can manipulate vocal fold vibratory patterns and vocal tract setup to maintain timbral consistency across an extended range.