Glottal characteristics of male speakers: Acoustic correlates and comparison with female data

Abstract

Acoustic measurements believed to reflect glottal characteristics were made on recordings collected from 21 male speakers. The waveforms and spectra of three nonhigh vowels (/ae, lambda, epsilon/) were analyzed to obtain acoustic parameters related to first-formant bandwidth, open quotient, spectral tilt, and aspiration noise. Comparisons were made with previous results obtained for 22 female speakers [H. M. Hanson, J. Acoust. Soc. Am. 101, 466-481 (1997)]. While there is considerable overlap across gender, the male data show lower average values and less interspeaker variation for all measures. In particular, the amplitude of the first harmonic relative to that of the third formant is 9.6 dB lower for the male speakers than for the female speakers, suggesting that spectral tilt is an especially significant parameter for differentiating male and female speech. These findings are consistent with fiberscopic studies which have shown that males tend to have a more complete glottal closure, leading to less energy loss at the glottis and less spectral tilt. Observations of the speech waveforms and spectra suggest the presence of a second glottal excitation within a glottal period for some of the male speakers. Possible causes and acoustic consequences of these second excitations are discussed.

Duration and spectral balance of intervocalic consonants: A case for efficient communication

“…The spectral slope of a speech segment is determined by the underlying articulation and by the syllable stress and is itself a perceptual cue for syllable stress (Sluijter, 1995a, b;Sluijter and Van Heuven, 1996;Sluijter et al, 1997;cf., Hanson, 1997;Hanson and Chuang, 1999;Tabain, 2003). However, it is difficult to measure the spectral slope in a uniform manner for different types of phonemes.…”

“…In the present study, the Spectral Center of Gravity was chosen as a measure of the spectral balance of the realizations and thus as a 'surrogate' measure of the spectral slope. The center of gravity of a spectrum is proportional to the air-speed/area in the constriction of obstruents for turbulent noise and is related to the speed of the vocal folds at the time of closure for sonorants (actually, the steepness of the air-flow profile at closure, see Hanson, 1997;Hanson and Chuang, 1999 for a detailed articulatory modeling of the spectral slope of the glottal source). That is, to increase the CoG of a fricative, ie., turbulence caused by a small opening, a speaker must either increase the lung pressure to increase the air-speed or decrease the size of the opening against the air pressure.…”

“…For the CoG it is known how it reacts to relevant acoustic parameters, like air-flow and turbulence, which are linked to "articulation strength" (cf., Boersma, 1998 section 4.3; cf., Hanson, 1997;Hanson and Chuang, 1999). In the present study, the Spectral Center of Gravity was chosen as a measure of the spectral balance of the realizations and thus as a 'surrogate' measure of the spectral slope.…”

“…Second, we needed an acoustic measure that can be used to link articulatory 'strength' (articulatory emphasis as used eg., by Fougeron and Keating, 1997) to the acoustics of speech over comprehensive phoneme inventories (cf., Chennoukh et al, 1997 for an alternative modeling effort; see also Jongman et al, 2000). The latter problem can be pragmatically solved by using the spectral slope as studied by Sluijter and others (Sluijter 1995(Sluijter a, 1995bSluijter and Van Heuven, 1996;Sluijter et al 1997;Hanson, 1997;Hanson and Chuang, 1999;Tabain, 2003), or the associated spectral balance, which is an indicator of the acoustic power relative to a normal speaking level (Boersma, 1998, section 4.3;cf., Hanson, 1997;Hanson and Chuang, 1999). Within any phone class, a flatter slope generally indicates perceptually louder speech (eg., Sluijter 1995 a, b;Sluijter and Van Heuven, 1996;Sluijter et al 1997) and may indicate more effort on the part of the speaker, eg., higher lung pressure and muscle tension (see section 2.2).…”

Začetne opombe k laringalizaciji v slovenščini

“…1996, Hanson 1997, Blomgren idr. 1998, Hanson in Chuang 1999, Švec idr. 2000, Berry 2001, Gerratt in Kreiman 2001, Hanson idr.…”