The Journal of the Acoustical Society of America · 1997 · 371 citations · 18 references
Language StudiesLinguisticsSpeech-language PathologyAcousticsPhoneticsPhonologyAcoustic MeasurementAcoustic PhoneticsExperimental PhoneticsPhonetic VariationSpeech AcousticsSuprasegmental PhoneticsFemale SpeechFormant StructureGender-related Speaker VariationLaryngeal AcousticsLaryngeal PhoneticsSpeech Signal AnalysisVariation In Voice QualityVoice Source Analysis
Theoretical and experimental evidence indicates that an open glottal configuration produces a volume‑velocity waveform with stronger low‑frequency components, weaker high‑frequency content, increased aspiration noise, and broader vocal‑tract bandwidths, especially of the first formant, all of which can be extracted from the speech spectrum. The study aims to formulate acoustic parameters of the voicing source that capture individual differences in female speakers’ voice qualities and to enhance the analysis and synthesis of female speech. These parameters were measured from the speech of a group of female speakers, and the authors hypothesized the corresponding glottal configurations. The authors identified a set of acoustic parameters that likely indicate glottal characteristics and suggested potential clinical applications.
The aim of the research reported in this paper is to formulate a set of acoustic parameters of the voicing source that reflect individual differences in the voice qualities of female speakers. Theoretical analysis and observations of experimental data suggest that a more open glottal configuration results in a glottal volume-velocity waveform with relatively greater low-frequency and weaker high-frequency components, compared to a waveform produced with a more adducted glottal configuration. The more open glottal configuration also leads to a greater source of aspiration noise and larger bandwidths of the natural frequencies of the vocal tract, particularly the first formant. These different attributes of the glottal waveform can be measured directly from the speech spectrum or waveform. A set of acoustic parameters that are likely to indicate glottal characteristics is described. These parameters are measured in the speech of a group of female speakers, and the glottal configurations of the speakers are hypothesized. This research contributes to the description of normal variations of voicing characteristics across speakers and to a continuing effort to improve the analysis and synthesis of female speech. It may also have applications in clinical settings.
18
Eva B. Holmberg, Robert E. Hillman, Joseph S. Perkell · The Journal of the Acoustical Society of America · 1988 · 636 citations