readings/kreimansidtis/media/1/laver/| Domain | |||
|---|---|---|---|
| Production/acoustics | fundamental frequency | voice | creak(y) |
| Perception | pitch | voice quality | creaky voice |
Comparison to percept of pitch and loudness
However, quality is multidimensional. It cannot be successfully scaled unidimensionally; and because more than one possible cue to quality exists, the possibility of listener differences is always present, so that quality can never have fixed acoustic determinants. Given this fact, the perceptual response evoked by a voice signal will always depend on factors like task demands, and listener attention will vary across the multiple facets of the signal, so that some are more important than others from occasion to occasion (although experimental controls can minimize the effects of these factors, as discussed below.). For this reason, a single perceived quality may not consistently result from a given signal, relative to the listener. In contrast, pitch and loudness do not ordinarily vary in this way, because of their more-or-less unidimensional nature. (page 9)
Signal/listener interaction
Voice quality is the result of perceptual processes, and must be defined in terms of both signals and listeners (page 10)
Methods of observation
to study voice in its narrow sense, most researchers adopt the more practical expedient of controlling for all non-laryngeal contributions to the sounds a speaker makes by restricting voice samples to steady state vowels (usually /a/). (page 6)
Physiological/acoustic/perceptual links
Issues of quality measurement have implications beyond the study of quality itself. Once the relationship between a signal and a percept is understood, it may be possibnle to determine which physiological parameters create perceptually meaningful changes in phonation. At present, it is not possible to determine which aspects of vocal physiology are perceptually important, in part because the relationship between perception and acoustics is poorly understood. (page 23)