Volume 53, Issue 1, January 1973

Electric network effects in the cochlea
View Description Hide DescriptionCochlear microphonics have been explained in terms of a varying resistance at the hair cells that modulates resting currents and voltages in an electric network powered by biological dc “batteries.” In this paper, the distributed electric circuit of a 2‐mm‐long section of the first turn of the cochlea is approximated by a lumped network consisting only of linear elements, and including both fixed and variable capacitances across membranes as well as the variable resistance. When this circuit is driven by sinusoidal parameter variations, the voltage and current analogous to the cochlear microphonic and its concomitant hair‐cell current exhibit these steady‐state phenomena: (1) frequency‐sensitive gain; (2) frequency‐ and amplitude‐sensitive phase shift; (3) dc shifts of both polarities in response to purely ac stimulation, analogous to the summating potential; (4) waveform distortion. These results show that an additional stage of signal processing may exist between basilar‐membrane vibration and nerve spike trains, and certainly between basilar‐membrane vibration and observed cochlear potentials.

Discrimination of fundamental frequency contours in synthetic speech: implications for models of pitch perception
View Description Hide DescriptionThe just‐noticeable difference (JND) for selected aspects of voice fundamental frequency (F _{0}) contours was determined by varying the F _{0} control parameter of a digitally simulated terminal analog speechsynthesizer. Data were obtained from three subjects for a number of 250‐msec segments of the synthetic vowel /ɛ/ differing only in fundamental frequency. Results indicate that the subjects can detect a change of 0.3 Hz in a constant F _{0} contour when F _{0} = 120 Hz, but the JND is an order of magnitude larger (2.0 Hz) when the F _{0} contour is a linear descending ramp (32 Hz/sec). Sensitivity to rate of change of F _{0} in linear ramps is surprisingly good; greatest sensitivity occurs when one ramp increases and the other decreases (JND = 12 Hz/sec). High‐pass filtering of the stimuli improves performance slightly, suggesting that the fundamental component is not involved in the detection of changes in F _{0}. Substitution of the synthetic stimulus /ya/ with its dynamic formant contours in place of /ɛ/ degrades performance only very slightly. Implications of these data for models of pitch perception mechanisms are discussed.

Auditory detection of frequency transition
View Description Hide DescriptionThe thresholds for the detection of frequency transition were obtained for the tone bursts in which the frequency was linearly changed from the initial value of 1000 Hz either during almost the whole burst duration or during a portion of it. The onset of transient had various delays from the beginning of the tone burst. The burst durations were 32, 102, 302, and 1000 msec. It was found that the nearly constant thresholds were obtained for a given burst duration as long as was held constant, where T _{1} was the duration of the initial segment of steady frequency and T _{2} was the duration of frequency transition. The thresholds were greater when the transition occurred near the burst onset than when it occurred near the cutoff, and in the former case, the thresholds were influenced by a rise time of tone burst. The thresholds were smallest when the frequency transitions occurred in the middle of the tone burst. It is suggested that there are two mechanisms for auditory detection of frequency transition: one is based on the pitch difference and the other on the gliding pitch.

The bowed string and the player
View Description Hide DescriptionRelations between bowing parameters, i.e., force applied, bow position, and velocity, are derived in terms of load impedance presented by bridge to string, characteristic impedance, and frictional coefficients. The range between least applied force needed to couple bow to load during sticking and maximum permitting uncoupling following sticking provides the generous tolerance, variable but typically in the ratio of 1 to 10, that makes the bowed string so flexible in performance. Domains of string behavior recognized by players are related graphically to bowing parameters. An electromagnetic method for observing particle velocity in the string reveals small but significant ripples caused by forces at the bow that the idealized explanation ignores. In one very flexible string, force exerted on the bridge varied approximately inversely with frequency out to the 15th harmonic, whereas for a string equivalent to a gut G for violin force became zero near the 7th. Elastic effects are considered, and it is suggested that in strings, either solid or wound, inharmonicity of 0.1 cent per square of mode number will not perceptibly degrade bowed‐string performance.

The influence of human factors on the performance of a real‐time speech recognition system
View Description Hide DescriptionThe performance of real‐time speech recognition systems is observed to depend upon the environments in which they are used. Human factors, caused for example by operator frustration, significantly affect success rates. A set of experiments with a word recognition system is described which quantifies this change, whilst the system is being used to control the movement of a displayed cross on a screen.

Shift of ear superiority in dichotic listening to temporally patterned nonverbal stimuli
View Description Hide DescriptionThe assumption that the direction of ear superiority in dichotic listening to sounds varies as a function of the number of stimulus transitions within sound sequences was tested in two experiments on 36 right‐handed students. In each sequence, three sounds which varied in terms of either frequency or duration were employed. Subject's task was to report the sequences by ear. The results showed that as the number of frequency or duration transitions increased from zero to two, ear superiority shifted from left to right. The shift from left to right ear superiority as a function of the increase in the complexity of temporal patterning suggests that perception of temporal patterns might be one of the underlying mechanisms in speech perception.

Perception of dichotically presented vowels
View Description Hide DescriptionListeners heard vowels in syllabic contexts, presented dichotically at S/N (signal‐to‐noise) ratios of 0 and −10 dB, and their responses were analyzed in terms of individual ears. In the most‐adverse listening condition the level of correct response in the right ear was 6% higher, on the average, than that obtained in the left ear; there was no difference between ears in the more‐favorable listening condition.

Acoustically augmented diffusional transport
View Description Hide DescriptionThe present work considers enhancement of mass transport resulting from superimposing an ultrasonically induced convective transport on a diffusional transport. As a result of the nonlinearities in the Navier‐Stokes equation, a time‐dependent secondary flow called acoustic streaming can be produced when an acoustic wave is passed through a medium. Between adjacent vortices or cells, molecular diffusion is the only means of transport; however, within each cell, mass transport is primarily by convection. Increases in the rate of mass transfer of the order of 150% above the normal diffusional flux were found. The theoretical analysis and results presented have application to systems in which ultrasonics may be used to increase mass transfer through membranes (e.g., dialysis), and to increase the efficiency of very active catalytic systems and of solid‐liquid extraction systems.

Propagation of elastic waves bound to a fluid layer between two solids
View Description Hide DescriptionTheoretical calculations and experimental measurements for an acoustic wave bound to and guided by a fluid layer between two semi‐infinite isotropic elastic solids are reported. Discussion of theory includes wave propagationcharacteristics, mode existence criteria, and the distribution of elastic energy in the structure, treating separately the cases where the geometry is symmetric and nonsymmetric about a plane of bisection. Experimental results include measured propagation characteristics for the layered geometry and a device realization of this particular structure.

Effect of diffusion on the growth and decay of acceleration waves in gases
View Description Hide DescriptionAs is well known, a small compressive disturbance in an elastic gas grows into a shock wave in finite time. If the gas is viscoelastic or has internal degrees of freedom, this result is no longer true unless the initial amplitude of the disturbance is larger than a critical amplitude. This article investigates the effect of diffusion on the growth and decay properties of small disturbances in a mixture of elastic gases. It is shown that a critical amplitude exists for the mixture. Thus, diffusion can be thought of as having a stabilizing effect in that not all compressive disturbances will grow into shock waves.

Three‐dimensional acoustic‐ray tracing in an inhomogeneous anisotropic atmosphere using Hamilton's equations
View Description Hide DescriptionThree‐dimensional ray‐tracing equations are presented for tracing acoustic rays in an inhomogeneousanisotropic atmosphere which are based on Hamilton's equations for geometrical optics. The equations are converted to a form suitable for direct numerical integration. Simplified ray equations for the special case of horizontal stratification are obtained. Results of ray tracing in an atmosphere with both vertical and angular wind shear are outlined, and the formation of a shadow zone in the downwind direction is demonstrated.

Acoustic radiation from a long solid rod into a semiinfinite liquid medium
View Description Hide DescriptionAn approximate numerical technique is presented whereby the pressure‐radiation patterns in a liquid half‐space produced by radiation from the end face of a circular cylindrical, semiinfinite, solid rod set flush in an infinite rigid baffle at the boundary of the half‐space can be calculated as a function of the elastic properties of the solid rod. The method is restricted to the low‐frequency range such that the lowest numbered mode in the solid rod is the only propagating mode. Axisymmetric longitudinal modes are the only ones assumed to exist in the rod. The velocity distribution of the rod end face does not have to be known or assumed. The radiation impedance of the rod end is calculated and compared with that of the rigid piston, as are the pressure radiation fields. The work here presents farfield radiation only, though the method can easily be extended to study the nearfield radiation patterns in the liquid half‐space.

Multiple scattering of sound by a periodic line of obstacles
View Description Hide DescriptionEquations derived earlier for multiple scattering by arbitrary three‐dimensional configurations are applied to a line of equally spaced identical obstacles. We develop several different representations and approximations for the field and derive the appropriate energy and scatteringtheorems. Spherically symmetric scatterers are considered as a special case. In general, the results are analogous to those obtained before for the two‐dimensional problem of scattering by a planar grating of identical cylinders but differ essentially in that now the scattered modes are conical instead of planar. A major difference is that resonance maxima (analogous to the Wood's anomalies for the planar grating) may occur not only for wavelengths that are slightly larger than the grazing (Rayleigh) wavelengths, but also for wavelengths that are slightly smaller. We also derive closed‐form approximations for small scatterers with arbitrary spacing, and then specialize the results to small spacing.

Optimized design: characteristic vibration shapes and resonators
View Description Hide DescriptionA technique is developed to find the axial mass distribution of rods and beams based upon desired properties of its first vibrational mode deflection function. This is an inverse Sturm‐Liouville problem for rods. The cases of particular interest are when a specified characteristic function does not satisfy the boundary conditions of the problem or continuity conditions. For example, a uniformly strained fixed‐free rod vibrating axially cannot exist because the strain must be zero at the free end. Using calculus of variations, a method is developed whereby the rod or beam is found whose first characteristic function can be as close as possible to the specified unattainable function. This is done for linearly elastic axially nonuniform materials vibrating axially or transversely. Examples include the application of the technique to the design of ultrasonicresonators.

Effect of transverse shear deformation on vibrations of planar structures composed of beam‐type elements
View Description Hide DescriptionA finite element beam‐type model considering transverse shear deformation is presented which utilizes Timoshenko's original kinematic variables. A direct calculation of the distribution of energies due to extension, flexure, and shear is indicated. For beam vibrations, comparison of the present results with other analyses shows that the current formulation is extremely effective for cross sections whose depth is comparable to a wavelength. Natural frequencies of frames and arches are computed using the present element and are compared with values based upon neglect of transverse shear deformation to illustrate this effect quantitatively.

Response of an elastic half‐plane to a moving and simultaneously fluctuating indenter
View Description Hide DescriptionBy means of integral‐transform techniques, the problem of steady motion and simultaneous harmonic oscillation of a smooth flat indenter is formulated and solved. The boundaries of the indenter are assumed to be fixed with respect to a coordinate system moving with steady velocity. A low‐frequency approximation for this case is obtained. The results of the approximation are subsequently used with a method of Crozier and Hunter to obtain the magnitude of the changing boundary when a curved or wedge indenter is considered.

Axial‐flexural vibration of frameworks using finite‐element approach
View Description Hide DescriptionA finite‐element procedure for predicting the frequency of axial‐flexural coupling vibration of frameworks is presented. In addition to the rectilinear beams, the formulation and the procedure are also shown to be highly accurate in the analysis of axial‐flexural coupling vibration of a portal frame by comparing the results with an analytical solution presented herein. Additional examples include a two‐bay portal frame and the single‐bay multi‐story frames. It is found that in the case where the horizontal beam contains floor masses or the frame is composed of multiple stories, substantial discrepancies exist between the axial‐flexural frequencies and the flexural frequencies. In those cases, the axial motion is no longer negligible in the vibrational analysis.

Some curious analytical ray paths for some interesting velocity profiles in geometrical acoustics
View Description Hide DescriptionGeometrical acoustics problems in oceanic applications are solved using complicated computer codes. This paper presents some cases which may be used to verify these codes. The velocity profiles are depth dependent, range dependent, and depth and range dependent. The ray path solutions, as well as the velocity profiles, are all expressed in terms of analytical, closed form, or at worst, tabulated functions.

Rytov's method and large fluctuations
View Description Hide DescriptionRytov's method has been the subject of arguments as to whether or not it has a greater domain of applicability than the method of small perturbations. In order to be able to use the result of Rytov's method one must assume that the sizes of the patches of inhomogeneity are much larger than the wavelength of the sound, i.e., λ≪l. Making use of this assumption, when we use the solution for Ψ′ to compare the discarded term ▿Ψ′^{2} with the remaining terms, we discover that Rytov's method has the same domain of applicability as the method of small perturbations. Using a simple geometrical model we deduce by physical reasoning the results of Mintzer and Bergmann. This model is then used to extend the theoretical prediction for the amplitude fluctuations in the (Mintzer) range to cover the case where the fluctuations are large.

Time‐varying multipath interference of broad‐band signals over a 7‐NM range in the Florida Straits
View Description Hide DescriptionThe frequency response function of a propagation channel is formulated by the coherent (vectorial) summation of CW signals arriving via many ray paths between source and receiver. Six months of continuous measurements of water temperature from a moored thermistor string in the Florida Straits are analyzed to determine the temporal variation of the sound speed profile. The observed average profile and its temporal variations are shown to be closely modeled by fitting the data with two layers of linear sound speed gradient. In order to investigate the effects on propagation of temporal variations of the two layers, a model of Refracted Bottom‐Reflected (RBR) propagation is developed for the particular case of a bottomed source and receiver. Model computations determine the relative influence of the parameters of the two layers upon the interference of broad‐band signals. The general features of multipath interference which are observed in the Florida Straits are presented and discussed. Model simulations of multipath interference are used to interpret experimental results. Some characteristics of the propagation channel (e.g., bandwidth of an interference fade) are shown to depend on average values of the model parameters, while others (e.g., CW amplitude) depend on small fluctuations about the averages and thus cannot be estimated with idealized propagation models. For typical profiles the gradient near the bottom determines the average properties of the channel frequency response function.