Analyzing Emotion Expression in Singing via Flow Glottograms, Long-Term-Average Spectra, and Expert Listener...

Area of Science:

• Acoustic phonetics
• Vocal pedagogy
• Music psychology

Background:

• Emotional expressivity in speech and singing is extensively studied.
• Emotional coloring is crucial in sung performance and vocal signaling.
• Physiological voice source parameters aid in understanding acoustic characteristics.

Purpose of the Study:

• To investigate the relationship between vocal emotional expression and acoustic parameters.
• To determine if voice source parameters (FLOGG) and long-term average spectrum (LTAS) correlate with perceived emotions in singing.
• To explore how acoustic features influence listener judgments of sung emotions.

Main Methods:

• Professional male singers performed scales in 10 emotional colors.
• Voice experts classified the sung emotions in a listening test.
• Acoustic analysis involved long-term average spectrum (LTAS) and flow glottograms (FLOGG) parameters.

Main Results:

• Emotions were grouped into four families: Anger-Contempt, Joy-Love-Pride, Calm-Tender-Neutral, and Sad-Fear.
• Listener recognition of emotion families exceeded chance levels.
• Vocal loudness significantly influenced LTAS and FLOGG parameters, with remaining parameters correlating and sorting into emotion-associated groups.

Conclusions:

• LTAS and FLOGG parameters significantly varied with singers' intended emotions.
• Voice source characteristics are reflected in LTAS parameters.
• LTAS parameters influence listener judgments of enacted emotions and accuracy of emotional coloring.