On the recognition of the Spanish fricatives /s/ and /f/

Abstract

The present work examines the perceptual load carried by the relative intensity of the noise to the intensity of the vowel in the identification of the Argentine Spanish fricatives /s/ and /f/. The estimuli consisted of synthetic fricative-vowel syllables. The fricative portion, consisting of a fixed band noise with a central frequency of 4500 Hz and a bandwidth of 300 Hz, was varied in its relative amplitude to the vowel, to obtain a set of ten noise amplitudes in relation to the fixed vowel amplitude. Spanish listeners identified low amplitude noises as /f/ and high ones as /s/. In a second experiment, subjects were tested on the same stimuli at two different overall sound-pressure levels of the syllable: 60 and 80 dB. For the 80-dB level, subjects identified more /f/ sounds than at the 60-dB level. This result can be interpreted as a different interaction between the vocalic portion and the fricative noise. It is suggested that this kind of interaction may be due to a differential masking effect of the vowel at different amplitudes.

The role of release bursts in the perception of [s]-stop clusters

“…While the preceding fricative noise may exert a slight masking effect on the burst, the amplitude of the following vocalic portion seems to have its perceptual effects primarily by changing the relative salience of cues contained in that portion itself. While the present data cannot be considered the last word on the issue, the possibility of a fixed perceptual criterion in the amplitude domain deserves further attention, both with regard to the perception of stop manner and to place-of-articulation distinctions in stops (see Ohde and Stevens, 1983)and fricatives (Gurlekian, 1981). The preceding experiments, experiment 2 in particular, demonstrate a remarkable sensitivity oflisteners to the presence of even very weak release bursts.…”

Recognition of Voiceless Fricatives by Normal and Hearing-Impaired Subjects

“…Jongman (1989) reported that shortening the frication noise to a duration of 70 ms does not significantly decrease fricative recognition performance compared to conditions where the full-duration cue is presented. The overall amplitude of the frication noise relative to the vowel also has been reported to be of some importance in distinguishing between different voiceless fricatives (Guerlekian, 1981;McCasland, 1979aMcCasland, , 1979band Stevens, 1985). But a recent study (Behrens & Blumstein, 1988) showed that the amplitude effect is small under more natural and controlled conditions where the spectral properties of the friction noise and the formant transitions are synthesized to be compatible.…”

Acoustic-phonetic features for the automatic classification of fricatives

“…Several researchers ͑Strevens, 1960; Heinz and Stevens, 1961;Guerlekian, 1981;Behrens and Blumstein, 1988a;Hedrick and Ohde, 1993;Borden et al, 1994͒ observed the difference in intensity between sibilants and nonsibilants and commented on the possibility of using the amplitude as a cue for such discrimination. Heinz and Stevens ͑1961͒ also found that nonsibilants were never correctly perceived by the listeners in their perceptual experiments, when the intensity was too high, no matter how the spectrum looked like.…”

“…Later studies ͑Fant, 1973; Cole and Cooper, 1975;McCasland, 1979a, b;Guerlekian, 1981;Manrique and Massone, 1981;Baum and Blumstein, 1987;Behrens and Blumstein, 1988a, b;Crystal and House, 1988;Jongman, 1989;Stevens et al, 1992;Hedrick and Ohde, 1993;Borden et al, 1994͒ have added to our knowledge about fricatives. However, except for a few studies, the acoustic characteristics that exist in the literature are qualitative, relational, and speaker dependent.…”