Comparison of different voice samples for perceptual analysis

Area of Science:

• Speech-Language Pathology
• Acoustic Analysis
• Perceptual Voice Assessment

Background:

• Perceptual judgments of voice quality are crucial in diagnosing voice disorders.
• The selection of voice sample material can significantly influence these judgments.
• Understanding the impact of different voice sample types is essential for accurate voice assessment.

Purpose of the Study:

• To validate the use of sustained vowels for perceptual voice analysis in French speakers.
• To investigate whether using only the stabilized portion of a sustained vowel underestimates dysphonia severity.

Main Methods:

• Recorded voice samples from 60 dysphonic patients and 20 controls.
• Collected three sample types: connected speech, complete sustained vowel, and stabilized sustained vowel.
• Utilized a jury of 7 experienced listeners for perceptual judgments and measured jury consistency.

Main Results:

• Jury consistency remained similar across all three voice sample materials.
• Perceptual judgments of dysphonia severity were less severe for stabilized sustained vowels compared to connected speech.
• Judgments based on complete sustained vowels were comparable to those based on connected speech.

Conclusions:

• Complete sustained vowels are suitable for perceptual voice analysis, yielding results comparable to connected speech.
• Stabilized sustained vowels alone may lead to underestimation of dysphonia severity.
• The choice of voice sample material is critical for accurate perceptual voice assessment.