Belting by Trained Singers Compared With Shouting by Untrained Nonsingers: How Do They Differ?

Objective: This exploratory study investigated differences between belting ("yell-like" singing) by trained singers and shouting by untrained participants. Special interest was in parameters estimating vocal efficiency and economy.

Methods: Ten trained singers with expertise in belting and 10 vocally untrained individuals (males and females) produced moderately loud and very loud syllable repetitions of /pe/ at three pitches (males: C#4, D#4, and F4 and females: F4, G4, and A4). Oral air pressure, flow, electroglottographic (EGG), and acoustic signals were recorded. Thirteen parameters were calculated, including subglottal pressure (P_sub), average air flow (Q), estimate of glottal resistance (GR), contact quotient (CQ), alpha ratio, level difference between the first and second harmonic (H1-H2), spectral center of gravity (CG), and sound pressure level (SPL). Twelve parameters estimating vocal efficiency and economy were calculated, including efficiency (VE), SPL/P_sub, SPL/CQ, and quasi-output-cost ratio (QOCR). Group differences between belters and shouters were explored separately for the males and females.

Results: Eighteen parameters in males and nine in females differed significantly (P < 0.05) between the groups. VE in dB (re 1 W) was higher in belting in both sexes. Male belters also showed lower Q, subglottal power (P_sub * Q), H1-H2, SPL/P_sub, and QOCR, as well as higher GR, CQ, and SPL/Q, than shouters. Female belters demonstrated lower P_sub and CQ, and higher CG, SPL/P_sub, QOCR, and SPL/CQ than shouters.

Conclusions: Trained belters demonstrated greater efficiency, producing similar acoustic output to the shouters, but with less subglottal power. This result may reflect differences in vocal skill. Somewhat counterintuitively, some economy estimates were lower in male belters than in shouters. This was due to the trained males' significantly higher P_sub and CQ. It is questionable whether the economy estimates approximating vocal fold impact stress based on CQ and P_sub perform well in high-intensity phonation.