Observer weighting of interaural delays in source and echo clicks.

A correlational analysis was used to assess the relative weight given to the interaural differences of time (IDTs) of source and echo clicks for echo delays ranging from 1-256 ms. In three different experimental conditions, listeners were instructed to discriminate the IDT of the source, the IDT of the echo, or the difference between the IDTs of the source and echo. The IDT of the target click was chosen randomly and independently from trial-to-trial from a Gaussian distribution (mu = 0 microsecond, sigma = 100 microseconds). The IDT of the nontarget click was either fixed at 0 microsecond or varied in the same manner as the IDT of the target. The data show that for echo delays of 8 ms or less, greater weight was given to the IDT of the source than to that of the echo in all experimental conditions. For echo delays from 16-64 ms, the IDT of the echo was weighted slightly more than that of the source and the weights accounted for a greater proportion of the responses when the echo was the target, indicating that the binaural information in the echo was dominant over the binaural information in the source. The data suggested the possibility that for echo delays from 8-32 ms, listeners were unable to resolve the temporal order of the source and echo IDTs. Listeners were able to weight the binaural information in the source and echo appropriately for a given task only when the echo delay was 128 ms or greater.

A simulated "cocktail party" with up to three sound sources.

Listeners identified spoken words, letters, and numbers and the spatial location of these utterances in three listening conditions as a function of the number of simultaneously presented utterances. The three listening conditions were a normal listening condition, in which the sounds were presented over seven possible loudspeakers to a listener seated in a sound-deadened listening room; a one-headphone listening condition, in which a single microphone that was placed in the listening room delivered the sounds to a single headphone worn by the listener in a remote room; and a stationary KEMAR listening condition, in which binaural recordings from an acoustic manikin placed in the listening room were delivered to a listener in the remote room. The listeners were presented one, two, or three simultaneous utterances. The results show that utterance identification was better in the normal listening condition than in the one-headphone condition, with the KEMAR listening condition yielding intermediate levels of performance. However, the differences between listening in the normal and in the one-headphone conditions were much smaller when two, rather than three, utterances were presented at a time. Localization performance was good for both the normal and the KEMAR listening conditions and at chance for the one-headphone condition. The results suggest that binaural processing is probably more important for solving the "cocktail party" problem when there are more than two concurrent sound sources.

The effect of distractor frequency on judgments of target laterality based on interaural delays.

A two-dimensional stimulus-classification paradigm was used to examine the ability of listeners to judge the laterality of an interaurally delayed low-frequency target component presented concurrently with a distractor component. Of primary interest was the effect on performance of the frequency difference (Delta f) between the target and distractor. In one set of conditions, the target was fixed at 753 Hz and the distractor was 353, 553, 653, 703, 803, 853, 953, or 1153 Hz (fixed within a block of trails). In a second set of conditions, the distractor was fixed at 753 Hz and the target frequency was 353, 553, 653, 703, 803, 953, or 1153 Hz. The listeners were presented with a target component with an interaural delay that varied from trial to trial, taking on one of ten values, five leading to the left ear and five leading to the right. A distractor component was simultaneously presented with an interaural delay that also took on one of the same ten values. Delays ranged from -90 to (+)90 microseconds in 20-microsecond steps. during a block of 100 trials, each of the possible combinations of target and distractor delay was presented once and only once in a random order. Listeners were instructed to make left-right judgments based on the target delay. Each condition was repeated ten times, and the slopes of the best linear boundaries between left and right responses were used to derive the relative weights given to the target and distractor. The duration of the signals was 200 microseconds. Two of the eight listeners weighted the target heavily when the target and distractor were spectrally remote but gave the two components equal weight when the different in frequency was small. These two listeners yielded similar target weights regardless of which component was designated as the target. One listener gave nearly equal weight to the target and the distractor regardless of Delta f. Five of the listeners gave greater weight to the higher of the two frequencies regardless of which was assigned as the target. This high-frequency dominance is explained in terms of cross-correlation functions based on the composite two-tone waveforms.

The synthetic-analytic listening task for modulated signals.

The synthetic-analytic listening task (SALT) developed by Dye and colleagues [Dye et al., J. Acoust. Soc. Am. 96, 2720-2731 (1994)] was applied to a task in which an amplitude-modulated tonal carrier was presented as a target during the standard stimulus. The standard stimulus was followed by a test stimulus in which the target along with another amplitude-modulated carrier (the distractor) was presented. The listener determined if in the test stimulus, the target (which was presented along with the distractor) was higher or lower in modulation depth than when the target was presented alone as the standard stimulus. The target and distractor were either 1- or 4-kHz carriers modulated at one of ten depths of modulation during the test stimulus at modulation rates ranging from 4 to 64 Hz. SALT was used to estimate the relative weight listeners assigned to the target and distractor as a function of the difference between their modulation rates, both for target carrier frequencies above and for target carrier frequencies below the distractor carrier frequency. When the target and distractor were modulated at the same rate, the target and distractor weights were equal, indicating synthetic listening. When the target and distractor differed in modulation rate, the listener gave more weight to the target suggesting a form of analytic listening. The result demonstrate the applicability of SALT to studies of modulation and reinforce the claim that different spectral components modulated with the same modulation pattern are processed synthetically.

Stimulus classification procedure for assessing the extent to which binaural processing is spectrally analytic or synthetic.

A two-dimensional stimulus classification paradigm was used to assess the extent to which listeners' processing of interaural delays at low frequencies is spectrally analytic or synthetic. Listeners were presented with a 753-Hz target with an interaural delay that varied from trial to trial, taking on one of ten values, five leading to the left ear and five leading to the right. A 553-Hz distractor component was simultaneously presented, with its interaural delay also presented at one of ten different values. During a block of 100 trials, each of the possible combinations of target and distractor delay was presented once, and only once, in a random order. Listeners were instructed to make left-right judgments based on the target delay. Each condition was repeated ten times, and the slopes of the best linear boundaries between left and right responses were used to derive the relative weights given to the target and distractor in judgments of laterality. Six of the nine listeners gave increasing weight to the target as the duration of the signals was increased from 25 or 50 to 400 ms. Three listeners showed little change with duration; one consistently gave equal weight to the target and distractor, two consistently gave greater weight to the target than to the distractor. The utility of classification paradigms in the study of multidimensional acoustic signals is discussed.

Discrimination of interaural envelope delays: the effect of randomizing component starting phase.

The goal of this study was to examine the nature of envelope extraction in the discrimination of high-frequency waveforms on the basis of envelope delay. Threshold interaural envelope delays were measured for complexes consisting of three or five components for which the starting phases of all sinusoids were either sine phase or randomized between intervals of a two-alternative forced-choice (2-AFC) task. The center frequency was 4 kHz and the frequency separation was varied from 25 to 500 Hz. The results showed that thresholds were greater for the phase-randomized conditions than the sine-phase conditions. The phase effect tended to diminish with increasing frequency separation for three-component complexes but not for the five-component complexes. Sensitivity to envelope delay was better for five-component complexes than for three-component complexes at most frequency separations. In general, the results showed superior lateralization performance for conditions in which the envelope fluctuations were greater, a finding that is consistent with models of high-frequency binaural processing that include envelope extraction prior to binaural comparison.

The combination of interaural information across frequencies: the effects of number and spacing of components, onset asynchrony, and harmonicity.

Threshold interaural delays were measured for a single interaurally delayed low-frequency target component presented against a background of two, four, six, or eight diotic "distractor" components. In the first experiment, a 753-Hz target and the flanking distractor components were gated on and off simultaneously. In subsequent experiments, the distractors were gated on 25-200 ms prior to the target. In addition, the target and distractor components were given various harmonic configurations. In general, threshold interaural delays were higher in all conditions in which distractors were present relative to thresholds obtained for the target component in isolation. Subjects reported that the pitch of the target component was more salient when an onset asynchrony between the target and distractors was present, but the components were perceived as occupying a single intracranial position in spite of the various interaural delays across the frequency domain. These results suggest that binaural processing of stimuli consisting of a small number of low-frequency temporally overlapping components occurs in a spectrally synthetic manner in which interaural information is combined across the spectrum, even in situations in which the segregation of pitch information occurs.

The combination of interaural information across frequencies: lateralization on the basis of interaural delay.

Three experiments were carried out that employed low-frequency tone complexes with interaural delays that varied across the frequency domain. In the first experiment, threshold interaural delays were measured for three-tone complexes for which one, two, or all three components were delayed. The center frequency was 750 Hz and the frequency spacing (delta f) between components was 20, 50, 100, 250, or 450 Hz. For all delta f's, the presence of two diotic components elevated the threshold interaural delays obtained for the third component relative to that obtained for a pure tone of the same frequency. In the second experiment, observers made left-right judgments regarding the direction of movement of signals for which two components were delayed by 25 microseconds to the left ear during one interval and to the right ear during the other interval, while a third component of a variable time difference was delayed to the opposite side as the tone pair. Subjects reported single intracranial images during each interval, and the data showed that interaural delays of one component to one ear could be offset by interaural delays of the other two components to the other ear. In the final experiment, threshold interaural delays were measured for five-tone complexes in which one, two, three, four, or five components were delayed. The center frequency was 750 Hz and delta f was fixed at 100 Hz. Thresholds decreased in a linear fashion as the number of delayed components increased, falling by about a factor of 5 as the number of delayed components went from one to five. These results are consistent with spectrally synthetic binaural processing, with the lateral position of intracranial images determined by a combination of interaural information across the spectrum. These effects could be brought about by a linear combination of the outputs of frequency-specific cross-correlation networks or by a wideband cross correlation of the signals at the two ears.

The combination of interaural time and intensity in the lateralization of high-frequency complex signals.

In an effort to examine the rules by which information arising from interaural differences of time (IDT) and interaural differences of intensity (IDI) is combined, d"s were measured for trains of high-frequency clicks (4000 Hz, bandpass) possessing various combinations of IDT and IDI. The number of clicks was either 1 or 8, with the interclick interval either 2 or 10 ms. A 2-IFC task was employed in which the paired values of IDT and IDI favored one side during one interval and the other side during the other interval. Data obtained with the combined cues are compared to those obtained with IDTs or IDIs alone in order to determine the degree to which processing of the two cues is done independently. Results suggest that lateralization with such stimuli is based on the sum of the temporal and intensive differences and not on independent evaluations of their separate presences.

Statistical and receiver operating characteristic analysis of empirical spike-count distributions: quantifying the ability of cochlear nucleus units to signal intensity changes.

Analytical methods from signal detection theory were applied in an effort to quantify the ability of cochlear nucleus (CN) units to signal changes in intensity. Of particular interest was the relation between this ability and the different patterns of discharge that characterize auditory neurons. Single-unit responses to best-frequency (BF) tone bursts were recorded from neurons in the gerbil cochlear nucleus, and empirical spike-count distributions were generated. The mean-to-variance ratios for regular units were generally larger than those of irregular units. Receiver operating characteristic (ROC) curves were generated from empirical spike-count distributions. The area under the ROC curve [P(A)] was computed and used to define the performance of an observer detecting whether or not a change in firing rate has occurred, thus signaling a change in intensity. For a given change in mean spike count, units characterized by regular interspike-interval (ISI) histograms typically gave larger P(A) values than did units characterized by irregular ISI histograms. In addition, onset units gave larger values of P(A) than did irregular units for a given change in mean spike count. These results suggest that regular and onset units are better able to signal intensity changes than are irregular units.

Discrimination of interaural differences of level as a function of frequency.

Discrimination of interaural differences of level (IDLs) was measured for pure tones as a function of frequency and as a function of the interaural difference of phase or level of a standard. Varying the interaural difference of the standard was assumed to change the lateral position of its intracranial image. Threshold IDLs were approximately constant over a frequency range from 200-5000 Hz, except in a region near 1000 Hz where they were slightly elevated. Thresholds increased as the value of the standard interaural differences of phase or level increased, implying that interaural resolution declines as the lateral image moves away from midline. The results are generally consistent with the predictions of current models of lateralization, but additions to these models are required in order for them to account for the slight frequency dependence of threshold IDLs.

Masking-level differences for trains of clicks.

Masking-level differences (MLDs) were measured for trains of 2000-Hz bandpass clicks as a function of the interclick interval (ICI) and the number of clicks in the train. The magnitude of the MLD grew as the number of clicks in the train was increased from 1 to 32. While the MLDs tended to be larger at longer ICIs, the effect was mediated by changes in detectability in the homophasic conditions. For click trains consisting of 4-32 clicks, the improvement in detectability in the antiphasic conditions with increases in the number of clicks appears to be the result of integration of acoustic power, as is the case for the homophasic conditions. The absence of MLDs for short trains of high-frequency transients remains quite puzzling, since large MLDs are found with single, low-frequency transients.

The effects of intensity on the detection of interaural differences of time in high-frequency trains of clicks.

Threshold values of interaural differences of time (delta IDTs ) were measured for trains of dichotic clicks whose levels were 20, 40, or 60 dB SPL. All clicks were bandpass filtered at 4 kHz, and the number of clicks in the train (n) was 1, 2, 4, 8, 16, or 32. The interclick interval (ICI) was 5, 2, or 1 ms. Performance was compared to that of an ideal integrator of information, which produces slopes of - 0.5 when log delta IDT versus log n is plotted. The results showed that increases in level had no effect on the slopes of the log-log functions regardless of the ICI but did decrease the intercepts. Shortening the ICI caused the slopes to go from nearly - 0.5 towards 0.0. The improvement with level could be explained by either a decrease in the temporal variability of neural discharges, or by an increase in the number of samples of IDT at higher intensities brought on by increased firing rates or the activation of more auditory units. A review of the physiological literature found the most parsimonious explanation to be that the decline in threshold IDT was mediated by an increase in the number of active units, each possessing the same degree of adaptation.

Detection of interaural differences of intensity in trains of high-frequency clicks as a function of interclick interval and number.

Listeners were asked to detect interaural differences of intensity in trains of 4000-Hz clicks as the interclick interval (ICI) was varied from 10 to 1 ms and the number of clicks in a train (n) was varied from 1 to 32. As has previously been shown for differences of time [Hafter and Dye, J. Acoust. Soc. Am. 73, 644-651 (1983)], plots of log interaural threshold versus log n produced slopes that decrease with ICI. These results are explained in terms of a saturation model which argues that as the click rate increases, the evoked neural activity changes from what is essentially a tonic response toward one that is more phasic.

Detection of interaural differences of time in trains of high-frequency clicks as a function of interclick interval and number.

Listeners were asked to detect interaural differences of time in trains of 4000-Hz clicks as the interclick interval (ICI) was varied from 10 to 1 ms and the number of clicks in a train (n) was varied from 1 to 32. Plots of log interaural threshold versus log n produce straight lines whose absolute slopes decrease toward 0.0 with decreasing ICI. These results are shown to fit a saturation model which argues that as the click rate increases, the evoked neural activity moves from a response that is tonic toward one which is more phasic. The need to postulate neural compression is based in part on the fact that the three most commonly cited models of the limitations imposed by high frequency--reduction in the depth of modulations due to narrow-band filtering within the auditory system, neural refractoriness, and nonindependence of successive samples of internal noise--do not predict a change in slope with rate.

Just-noticeable differences of frequency for masked tones.

Frequency discrimination was measured as a function of level for tones of 500, 1000, 2000, 3000, and 4000 Hz. The tones were masked by a wide-band noise whose level was set to maintain constant E/No. Frequency discrimination was found to improve with level for tones of 500 and 1000 Hz while it grew worse for tones of 3000 and 4000 Hz. The data for 500 and 1000 Hz are accounted for by the periodicity-type neural timing model of Green and Luce ["Counting and Timing Mechanisms in Auditory Discrimination and Reaction Time," in Contemporary Developments in Mathematical Psychology, edited by D. H. Krantz, R. C. Atkinson, R. D. Luce, and P. Suppres (Freeman, San Francisco, 1974). Vol. 2]. The data for 3000 and 4000 Hz are explained by the modified energy detection model of Henning [J. Acoust. Soc. Am. 42, 1325-1334 (1967b)].

Lateralization of tonal signals which have neither onsets nor offsets.

In order to ascertain the special importance of binaural cues conveyed in the transient portions of dichotic signals, thresholds for interaural differences of time (delta t) and intensity (delta I) were studied using stimuli whose onsets and offsets were masked. Intense noise was used to mask all portions of each experimental trial except for the two intervals of a two-interval, forced-choice detection task. During the intervals, the noise was turned off with decay-rise times of 10 ms. What remained were tones whose interaural phase or intensity was different for intervals one and two. Performance was compared to control conditions which used unmasked gated sinusoids. For longer durations, detection without onsets and offsets was about as good as that with no masker. For the shorter signals, detection without transients was poorer than with standard lateralization, but this is attributed to forward and backward masking which reduced the effective durations of those stimuli. The ability to detect interaural differences of time with the onsets and offsets masked was extended to conditions in which the decay times of the noise were 100 ms. Performance here was slightly worse, but not by so much as to change the basic result. This is interpreted as showing that performance with the faster decay-rise times was not a product of momentary undershoots in neural following, but depended, rather, upon a true encoding of the interaural information in the stimulus fine-structure.