Cue Interaction in the Perception of Prosodic Prominence: The Role of Voice Quality

Abstract

Voice quality is an important dimension in human communication, used to mark a variety of phenomena in speech, including prosodic prominence. Even though numerous studies have shown that speakers modify their voice quality parameters for marking prosodic prominence, the impact of these modifications on perceived prominence is less studied. Our investigation looks at the effect of a well-known measure of voice quality, cepstral peak prominence (CPP), on syllabic prominence ratings given by both naive and expert listeners. Employing read speech materials in German, we quantify the role of CPP alone and in combination with other acoustic cues marking prominence, namely intensity, duration and fundamental frequency. While CPP, by itself, had a significant effect on the perceived prominence for most of the listeners, when used in conjunction with the other cues, its impact was reduced. Moreover, when assessing the importance of each of these four cues for determining the perceived prominence score we found important individual variation, as well as differences between naive and expert listeners.

ha-HA-hha? Intensity and voice quality characteristics of laughter

“…Previous work has shown cpp to be the best signal-based correlate for perceived voice quality in continuous speech [25], being able to capture also changes in speech characteristics due to linguistic processes (e.g. prosodic prominence [26]).…”

Classification of voice quality using neck-surface acceleration: Comparison with glottal flow and radiated sound

“…12 In addition, voice quality has been increasingly regarded as a prosodic feature, related to such suprasegmental phenomena as declination, 13 prosodic boundaries 14 and prosodic prominence. 15 Given the relevance of voice quality in such diverse communicative and clinical contexts, much research has been done on automatic and instrumental discrimination between voice quality categories. Work on this topic has employed a wide range of measures across a multitude of domains, including temporal, 16 spectral 17−19 and cepstral features.…”

Two Pragmatic Functions of Breathy Voice in American English Conversation

“…In Spanish, Slovak, and English it also helps differentiate among functions of affirmative cue words, such as backchanneling, providing feedback or communicating agreement [15]. Breathy voice also tends to be characteristic of non-prominent syllables in German [16]. Our own previous work on American English identified two roles relevant to the functions reported below: a role for breathy voice in marking "interpolated, meta-and high-priority utterances" [17], and a role in indicating dissatisfaction [18], although follow-up work found that breathy voice over about 3 seconds, with a short halfsecond excursion to non-breathy voice, was a better indicator of the latter (J. Avila, personal communication).…”