A new approach to sound source segregation.

We rely critically on our ability to 'hear out' (segregate) individual sound sources in a mixture. Yet, despite its importance, little is known regarding this -ability. Perturbation analysis is a psychophysical method that has been successfully applied to related problems in vision [Murray, R.F. 2011. J. of Vision 11, 1-25]. Here the approach is adapted to audition. The application proceeds in three stages: First, simple speech and environmental sounds are synthesized according to a generative model of the sound--producing source. Second, listener decision strategy in segregating target from non--target (noise) sources is determined from decision weights (regression coefficients) relating listener judgments regarding the target to lawful perturbations in acoustic parameters, as dictated by the generative model. Third, factors limiting segregation are identified by comparing the obtained weights and residuals to those of a maximum-likelihood (ML) observer that optimizes segregation based on the equations of motion of the generating source. Here, the approach is applied to test between the two major models of sound source segregation; target enhancement versus noise cancellation. The results indicate a tendency of noise segregation to preempt target enhancement when the noise source is unchanging. However, the results also show individual differences in segregation strategy that are not evident in the measures of performance accuracy alone.

Discrimination of the spectral density of multitone complexes.

Spectral density (D), defined as the number of partials comprising a sound divided by its bandwidth, has been suggested as cue for the identification of the size and shape of sound sources. Few data are available, however, on the ability of listeners to discriminate differences in spectral density. In a cue-comparison, forced-choice procedure with feedback, three highly practiced listeners discriminated differences in the spectral density of multitone complexes varying in bandwidth (W = 500-1500 Hz), center frequency (f(c) = 500-2000 Hz), and number of tones (N = 6-31). To reduce extraneous cues for discrimination, the overall level of the complexes was roved, and the frequencies were drawn at random uniformly over a fixed bandwidth and center frequency for each presentation. Psychometric functions were obtained relating percent correct discrimination to ΔD in each condition. For D < 0.02 Hz(-1), the steepness of the functions remained constant across conditions, but for D > 0.02 Hz(-1), they increased with D. The increase, moreover, was accompanied by a reduction in the upper asymptote of the functions. The data were well fit by a model in which spectral density discrimination is determined by the frequency separation of components on an equivalent rectangular bandwidth scale, yielding a roughly constant Weber fraction of ΔD/D = 0.3.

Modeling manner of contact in the synthesis of impact sounds for perceptual research.

Impact sounds synthesized according to a physical model have increasingly become the stimulus of choice in studies of sound source perception. Few studies, however, have incorporated manner of contact in their models because of the complexity of the mechanics involved. Here a simplified model of contact is described suitable for application to perceptual research. The results of the simplified model are shown to be in good agreement with those of more comprehensive numerical methods receiving prior acoustic validation [Chaigne and Lambourg, J. Acoust. Soc. Am. 109, 1422-1432 (2001)]. The advantages of the model for applications to perceptual research are discussed.

Auditory discrimination of force of impact.

The auditory discrimination of force of impact was measured for three groups of listeners using sounds synthesized according to first-order equations of motion for the homogenous, isotropic bar [Morse and Ingard (1968). Theoretical Acoustics pp. 175-191]. The three groups were professional percussionists, nonmusicians, and individuals recruited from the general population without regard to musical background. In the two-interval, forced-choice procedure, listeners chose the sound corresponding to the greater force of impact as the length of the bar varied from one presentation to the next. From the equations of motion, a maximum-likelihood test for the task was determined to be of the form Δlog A + αΔ log f > 0, where A and f are the amplitude and frequency of any one partial and α = 0.5. Relative decision weights on Δ log f were obtained from the trial-by-trial responses of listeners and compared to α. Percussionists generally outperformed the other groups; however, the obtained decision weights of all listeners deviated significantly from α and showed variability within groups far in excess of the variability associated with replication. Providing correct feedback after each trial had little effect on the decision weights. The variability in these measures was comparable to that seen in studies involving the auditory discrimination of other source attributes.

Sensory constraints on auditory identification of the material and geometric properties of struck bars.

A computational formula is derived for estimating the constraints limited auditory sensitivity imposes on auditory identification of the material and geometric properties of struck bars. The formula combines a model of the transverse motion of the bar with empirical psychometric functions to map out "null" regions in the bar's physical parameter space where changes in the frequency, amplitude, and decay of partials are likely below threshold for detection. Parameters of the physical space include bar density, Young's modulus, fluid and viscoelastic damping factors, bar length, and bar cross-sectional area (as related to bar shape and hollowness). The formula is used to estimate the possible effect of limited sensitivity in past studies on the auditory identification of bar attributes. The results suggest that sensitivity may, indeed, have played a role in some studies, and that apparent discrepancies in results may be understood based on whether the predominant source of damping was internal or external to the bar. The formula identifies conditions representing an expected bound on identification performance and thereby may be used to aid in the design of future studies for which the struck bar is the sound source of choice.