Temporal integration of sinusoidal increments in the absence of absolute energy cues.

Classical temporal integration (TI) is often viewed as a frequency-dependent, energy-based detection process. Detection thresholds for brief sinusoidal increments in either a fixed-level or a random-level broadband pedestal are reported that refute this traditional perspective of TI, Instead, evidence is presented that indicates (a) detection of absolute energy is not necessary for the TI effect and (b) the frequency dependence of TI is consistent with variations across frequency in peripheral auditory tuning, rather than the integration process per se. When peripheral frequency selectivity is controlled, TI can be explained by a frequency-invariant integration process over at least the frequency range from 500 to 4,000 Hz. This process is characterized by threshold improvements of 8-9 dB per decade increase in duration for increment durations between 10 and 300 ms.

Rate and timing cues associated with the cochlear amplifier: level discrimination based on monaural cross-frequency coincidence detection.

The perceptual significance of the cochlear amplifier was evaluated by predicting level-discrimination performance based on stochastic auditory-nerve (AN) activity. Performance was calculated for three models of processing: the optimal all-information processor (based on discharge times), the optimal rate-place processor (based on discharge counts), and a monaural coincidence-based processor that uses a non-optimal combination of rate and temporal information. An analytical AN model included compressive magnitude and level-dependent-phase responses associated with the cochlear amplifier, and high-, medium-, and low-spontaneous-rate (SR) fibers with characteristic frequencies (CFs) spanning the AN population. The relative contributions of nonlinear magnitude and nonlinear phase responses to level encoding were compared by using four versions of the model, which included and excluded the nonlinear gain and phase responses in all possible combinations. Nonlinear basilar-membrane (BM) phase responses are robustly encoded in near-CF AN fibers at low frequencies. Strongly compressive BM responses at high frequencies near CF interact with the high thresholds of low-SR AN fibers to produce large dynamic ranges. Coincidence performance based on a narrow range of AN CFs was robust across a wide dynamic range at both low and high frequencies, and matched human performance levels. Coincidence performance based on all CFs demonstrated the "near-miss" to Weber's law at low frequencies and the high-frequency "mid-level bump." Monaural coincidence detection is a physiologically realistic mechanism that is extremely general in that it can utilize AN information (average-rate, synchrony, and nonlinear-phase cues) from all SR groups.

Evaluating auditory performance limits: i. one-parameter discrimination using a computational model for the auditory nerve.

A method for calculating psychophysical performance limits based on stochastic neural responses is introduced and compared to previous analytical methods for evaluating auditory discrimination of tone frequency and level. The method uses signal detection theory and a computational model for a population of auditory nerve (AN) fiber responses. The use of computational models allows predictions to be made over a wider parameter range and with more complete descriptions of AN responses than in analytical models. Performance based on AN discharge times (all-information) is compared to performance based only on discharge counts (rate-place). After the method is verified over the range of parameters for which previous analytical models are applicable, the parameter space is then extended. For example, a computational model of AN activity that extends to high frequencies is used to explore the common belief that rate-place information is responsible for frequency encoding at high frequencies due to the rolloff in AN phase locking above 2 kHz. This rolloff is thought to eliminate temporal information at high frequencies. Contrary to this belief, results of this analysis show that rate-place predictions for frequency discrimination are inconsistent with human performance in the dependence on frequency for high frequencies and that there is significant temporal information in the AN up to at least 10 kHz. In fact, the all-information predictions match the functional dependence of human performance on frequency, although optimal performance is much better than human performance. The use of computational AN models in this study provides new constraints on hypotheses of neural encoding of frequency in the auditory system; however, the method is limited to simple tasks with deterministic stimuli. A companion article in this issue ("Evaluating Auditory Performance Limits: II") describes an extension of this approach to more complex tasks that include random variation of one parameter, for example, random-level variation, which is often used in psychophysics to test neural encoding hypotheses.

Evaluating auditory performance limits: II. One-parameter discrimination with random-level variation.

Previous studies have combined analytical models of stochastic neural responses with signal detection theory (SDT) to predict psychophysical performance limits; however, these studies have typically been limited to simple models and simple psychophysical tasks. A companion article in this issue ("Evaluating Auditory Performance Limits: I") describes an extension of the SDT approach to allow the use of computational models that provide more accurate descriptions of neural responses. This article describes an extension to more complex psychophysical tasks. A general method is presented for evaluating psychophysical performance limits for discrimination tasks in which one stimulus parameter is randomly varied. Psychophysical experiments often randomly vary a single parameter in order to restrict the cues that are available to the subject. The method is demonstrated for the auditory task of random-level frequency discrimination using a computational auditory nerve (AN) model. Performance limits based on AN discharge times (all-information) are compared to performance limits based only on discharge counts (rate place). Both decision models are successful in predicting that random-level variation has no effect on performance in quiet, which is the typical result in psychophysical tasks with random-level variation. The distribution of information across the AN population provides insight into how different types of AN information can be used to avoid the influence of random-level variation. The rate-place model relies on comparisons between fibers above and below the tone frequency (i.e., the population response), while the all-information model does not require such across-fiber comparisons. Frequency discrimination with random-level variation in the presence of high-frequency noise is also simulated. No effect is predicted for all-information, consistent with the small effect in human performance; however, a large effect is predicted for rate-place in noise with random-level variation.

A phenomenological model for the responses of auditory-nerve fibers: I. Nonlinear tuning with compression and suppression.

A phenomenological model was developed to describe responses of high-spontaneous-rate auditory-nerve (AN) fibers, including several nonlinear response properties. Level-dependent gain (compression), bandwidth, and phase properties were implemented with a control path that varied the gain and bandwidth of tuning in the signal-path filter. By making the bandwidth of the control path broad with respect to the signal path, the wide frequency range of two-tone suppression was included. By making the control-path filter level dependent and tuned to a frequency slightly higher than the signal-path filter, other properties of two-tone suppression were also included. These properties included the asymmetrical growth of suppression above and below the characteristic frequency and the frequency offset of the suppression tuning curve with respect to the excitatory tuning curve. The implementation of this model represents a relatively simple phenomenological description of a single mechanism that underlies several important nonlinear response properties of AN fibers. The model provides a tool for studying the roles of these nonlinearities in the encoding of simple and complex sounds in the responses of populations of AN fibers.

Detection of time- and bandlimited increments and decrements in a random-level noise.

The purpose of this study was to compare detection of increments and decrements occurring over limited regions of time and frequency within a 500-ms broadband (0-6000 Hz) noise. Three listeners tracked detection thresholds adaptively in a two-interval, two-alternative forced-choice task. Thresholds were measured for both increments and decrements in level [delta L = 10 log10(1 + delta N0/N0) dB, where N0 is the spectral power density of the noise] as a function of signal duration (T = 30-500 ms) for a range of signal bandwidths (W = 62-6000 Hz) that were logarithmically centered around 2500 Hz. Listeners were forced to rely on temporal- and spectral-profile cues for detection due to randomization of overall presentation level from interval to interval, which rendered overall energy an inconsistent cue. Increments were detectable for all combinations of W and T, whereas decrements were not consistently detectable for W < 500 Hz. Narrow-band decrements were not detectable due to spread of excitation from the spectral edges of the noise into the decrements. Increment and decrement thresholds were similar for W > or = 1000 Hz. Temporal- and spectral-integration effects were observed for both increments and decrements. The exceptions were for random-level conditions in which the signal matched the bandwidth or duration of the standard. A multicue decision process is described qualitatively to explain how the combination of temporal- and spectral-profile cues can produce temporal- and spectral-integration effects in the absence of overall-energy cues.

Masked detection thresholds and temporal integration for noise band signals.

Masked detection thresholds were measured at a center frequency of 2500 Hz for a range of noise signal bandwidths (W = 62 to 6000 Hz) and durations (T = 10 to 480 ms) approximating that found acoustically in speech. The signals were presented in an uncorrelated 480-ms, 6000-Hz-wide masker. The masker was presented: (1) at a constant spectrum level (53 dB SPL/Hz) or (2) with the overall level varied randomly over a 50-dB range from interval to interval of a trial. Performance was disrupted in the random-level masker only for the condition where the signal and uncorrelated masker were gated on and off simultaneously and were matched spectrally. Time constants (tau) estimated from temporal integration functions fit to the masked detection threshold data were related inversely to W for W broader than the critical bandwidth. Sensitivity to the noise signals was evaluated in the context of an optimum-detection model (Green, 1960). The results did not follow the prediction of a constant bandwidth-duration (WT) product, but may be understood in terms of cues available to the listener from the relative combination of signal and masker parameters. At least three cues for detection were identified in these experiments: (1) a relative timing cue, (2) a spectral shape cue, and (3) a traditional energy cue compared across observation intervals. The relative timing cue and spectral shape cue together contributed as much as a 10- to 12-dB advantage relative to detection based on the traditional energy cue alone. A new multi-cue detection model for predicting the masked detection thresholds was derived. Predictions from the new model were highly correlated (r = 0.95) with the empirically measured masked detection thresholds.