Listener Experience and Perception of Voice Quality

Abstract

Five speech-language clinicians and 5 naive listeners rated the similarity of pairs of normal and dysphonic voices. Multidimensional scaling was used to determine the voice characteristics that were perceptually important for each voice set and listener group. Solution spaces were compared to determine if clinical experience affects perceptual strategies. Naive and expert listeners attended to different aspects of voice quality when judging the similarity of voices, for both normal and pathological voices. All naive listeners used similar perceptual strategies; however, individual clinicians differed substantially in the parameters they considered important when judging similarity. These differences were large enough to suggest that care must be taken when using data averaged across clinicians, because averaging obscures important aspects of an individual’s perceptual behavior.

Perception of aperiodicity in pathological voice

“…However, observed associations between acoustic and perceptual measures have varied considerably across studies. For example, correlations between measures of jitter and ratings of both breathiness and roughness have ranged from 0 to about 0.7 ͑for review see Heiberger and Horii, 1982;Kreiman and Gerratt, 2000͒. Multidimensional scaling studies ͑which examine the perceptual dimensions that underlie perceived vocal similarity͒ have also provided inconsistent results with respect to the role that jitter, shimmer, and the NSR play in determining quality ͑Kreiman et al, 1990;Kreiman and Gerratt, 1996͒. Such variability in results has undermined confidence in both the acoustic measures and their perceptual interpretation.…”

The multidimensional nature of pathologic vocal quality

“…Mildly, moderately, and severely breathy and rough voices were all represented, as were a variety of diagnoses. A previous multidimensional scaling study using these voices (Kreiman et al, 1990) revealed breathiness and roughness dimensions which each accounted for more than 25% of the variance in listeners' dissimilarity judgments. 2 Voice samples were low-pass filtered using two 4-pole Butterworth filters with cutoff frequencies of 6300 Hz, and two with cutoff frequencies of 7500 Hz, for a total reduction in amplitude of 3.2 dB at 5.6 kHz and 39.4 dB at 9 kHz.…”

“…Kreiman et al (1990) found dimensions correlated with rated breathiness and roughness in a MDS study of 18 pathological male voices. However, "rough" and "breathy" dimensions did not consistently emerge from a subsequent study examining individual differences in voice perception (Kreiman et al, 1992), suggesting these dimensions are not perceptually important for every listener, even in a fixed perceptual context.…”

Validity of rating scale measures of voice quality

“…One way to motivate scales for specific aspects of quality is with reference to overall quality: Individual scales or sets of scales may be valid to the extent that as a group they measure overall quality. For example, studies using multidimensional scaling ͑Murry et al., 1977;Kreiman et al, 1990Kreiman et al, , 1992Kreiman et al, , 1994Kempster et al, 1991;Kreiman and Gerratt, 1996͒ attempt to identify the perceptual dimensions that underlie listeners' judgments of the overall similarity of pairs of voices. Unfortunately, traditional scales have not generally emerged as perceptual dimensions from these studies, which in consequence provide little support for the validity of such scales as measures of overall vocal quality.…”