885095991

8:30

4PP3. Dichotic profile analysis. Gaił M. Whitelaw, Chien-Yeh Hsu, and Lawrence L. Feth (Div. of Speech and Hcaring Sci., Ohio State Univ._f 110 Pressey Hall, 1070 Carmack Rd., Columbus, OH 43210)

The majority of profile analysis studies have used monotic and diotic stimulus presentation. Use of dichotic stimulus presentation has been limited in profile analysis experiments, with the presentation configura-tion primarily consisting of the signal component presented to one ear and the nonsignal components presented to the opposite car [e.g., Bernstein and Green, J. Acoust. Soc. Am. 81, 1888-1895 (1987)). In generał, results obtained in profile analysis experiments using dichotic stimulus presentation have revealed listener performance inferior to that obtained by either monotic or diotic presentation. Profile analysis is investigated by comparing a monotic condition to a dichotic condition. The monotic signal was generated by producing 21 components using equal logarithmic spacing. Level per component was equal in the standard signal. To create the profile, the odd-numbered components were incrcmented and the even-numbered components were deeremented. The dichotic signal was created by presenting the even-numbered components to the opposite ear. A 2Q-2AFC roving level paradigm was used. Preliminary results suggest that listeners are unable to perform the profile analysis task when information is presented dichotically. Discus-sion will center on the implications of failing to obtain a dichotic profile.

8:45

4PP4. The roles of pitch difterences and modulation incoherence In concurrent sound segregatlon. Robert P. Carlyon (Exptl. Psychol., Sussex Univ., Brighton BN1 9QG, England), Laurent Demany, and Catherine Semal (Lab. de Psychoacoust., Univ. de Bordeaux, Bordeaux, France)

Listeners discriminated pairs of complex sounds, each consisting of two groups of components. Only those components in the lower-frequency group were resolvable by the peripheral auditory system. For the standard stimulus, the /0’s of the two groups were freąuency mod-ulated coherently at about 125 Hz, so that they were always equal. For the signal, the two /U’s were modulatcd incohcrently so that they dif-fered by an amount that oscillated between values proportional to the depth of FM. Listeners could perform the discrimination when the zero-peak FM depth was about 7%, and two findings indicate that they did so by simultaneously comparing the pitches of the two groups of components. First, listeners could do the task when the lower group consisted of only odd harmonics of 125 Hz (pitch = 125 Hz), but not when it consisted of even harmonics (pitch = 250 Hz). This shows that they were not comparing the ratę of amplitudę modulation, caused by beating between adjacent components in the lower group, to that in the upper group: These rates were the same in both conditions. Second, performance was at chance when the /0’s of the two groups differed (e.g., 100 and 225 Hz), indicating that listeners could not detect FM incoherence per se.

9*)0

4PP5. Psychoacoustic evaluatk>n of random-like deviation changes In FM signals. Edward Ozimek (Psychoacoust. Lab., Dept. of Psychol., Univ. of Florida, Gainesville, FL 32611)

The detection threshold for frequency modulation was measured at several carrier frequencies. The standard (unmodulated) waveform was a sinusoidal signal of fixed amplitudę and fixed frequency. The compar-ison waveform was frequency modulated. The modulating signal had constant frequency, but the amplitudę was a random variable. The amplitudę of each successivc period was the absolute value of a Gaussian random variable with mean, 0, and standard deviation, a. The ability of listeners to detect the FM modulation by adaptively varying o in a two-altemative forced-choice procedurę was determined. An inerease in the carrier frequency from 120 to 1000 Hz causes, on average, a threc-fold inerease in the threshold value of a. Changes in modulation fre-ąuencies from 4 to 16 Hz inerease the threshold by twofold at the highest carrier frequency. At low-modulation frequencies, different amplitudę samples produce statistically different discrimination thresholds. Compared with the classical data for FM signals, the random deviations are easier to hear at higher carrier frequcncies and harder to hear at lower carrier frequencies. [Research supported by the Kościuszko Foundation and the NIH.)

9:15

4PP6. Scalp potentials elicited using AM acoustlc signals. William F. Dolphin and David C. Mountain (Dept. of Biomedical Eng., Boston Univ., Boston, MA 02215)

Temporal information is one of the most fundamental aspects of an acoustic signal extracted by the auditory system. Especially as regards communication sounds, temporal patteming is an extremely important information carrying feature of complex signals. Modulation of signal amplitudę, at frequencies up to scveral hundred Hz, is common in many species specific vocalizations as well as in human speech. Scalp potentials which follow the low-frequency envelope of an ampłitude-modulated (AM) stimulus waveform were evoked and recorded from anesthetized gerbils. This envelope following response is presumably due to the synchronized discharge of populations of neurons in the auditory pathway. When the carrier frequency (/<.) of an AM constant-frequency acoustical stimulus is greaier than the electrical cutoff fre-quency of the inner hair cells and the modulating frequency (J_mod) is much below this cutoff, the instantancous firing ratę of the auditory nerve approximately follows the envelope of the stimulus. As the max-imum amplitudę of the basilar membranę response occurs at the region corresponding to/, it is hypothesized that inner hair cells in this region are preferentially stimulated. To test this hypothesis, AM stimuli in the presence of pure-tone maskers of varying frequencies were presented and scalp potentials recorded. A strong envelope following response, at the frequency of the modulator, was obtained to AM stimuli alone. The response was quite robust and could be elicited using a wide rangę of //s, modulation frequencies, and stimulus intensities. The response could be eliminated by the simultaneous presentation of a pure-tone masking stimulus (Z**). When /_nml was close to/, responses at the modulation frequency were diminished by up to 25 dB. Measurements of the group delay. determined from the phase of the response relative to the stimulus phase, indicate at least three separate brain regions as the site of th»s response. The responses come from progressively morę central regions as the modulation frequency is reduced. Such studies may yield valuable insights to the processing of complex stimuli by the auditory system.

9:30

4PP7. The role of envelope cues for the detection of a tonę added to a narrow band of noise. Virginia M. Richards (Dept. of Psychol., Univ. of Pennsylvania, 3815 Walnut St., Philadelphia, PA 19104) and Robert D. Nckrich (Univ. of Pennsylvania, Philadelphia, PA 19104)

Using a 21FC paradigm, psychometrie functions for the detcctability of a tonal signal added to a 40-Hz-wide band of Gaussian noise was measured using signal and center frequencies of 600,1800, and 5400 Hz. In a second condition, the noise-alonc and the touc-plus-noise stimuli were scaled so as to yield identical energy values across the two intervals of each trial. Undcr the assumption of optimal combination of independent cues, a comparison of these conditions allows an estimate of the relative contribution of energy-based and non-energy-based cues for the detection of a tonę added to noise. In a third condition, the envelopes of the scaled noise-alone and tone-plus-noise stimuli were extracted, and used to modulatc tones of either 600, 1800, or 5400 Hł The ability to discriminate between enve!ope patterns was measured, and that measurc used to estimate each observer’s reliance on envelope-based cues. Large individual difTerences were obtained. For one observer, the detection of a tonę added to Gaussian noise was determined to rely equally on envelope-based and energy-based cues. A second obsenrer was found to

1912

J. Acoust. Soc. Am., Vol. 89, No. 4, Pt. 2, April 1991

121st Meeting: Acoustical Sodety of America

1912