Volume 36, Issue 6, June 1964

Distortion of the Temporal Pattern of Speech: Interruption and Alternation
View Description Hide DescriptionIn 1953, Cherry found that a subject can repeat virtually all of a continuous message that is switched alternately to his left and right ears, except at certain critical rates of alternation around 3 cps. The experiments reported here investigated the effect described by Cherry. In the first experiment, the intelligibility of continuous speech was measured as a function of rate of alternation, which ranged from 1–16 cps. Two such functions were measured, one using normal speech and the other using the same speech played back at slightly increased speed. Comparison of the two functions shows that, when the playback speed of the speech was increased, the rate of alternation that gave subjects most difficulty also increased by the same factor. Therefore, the effect cannot be ascribed to any processing time, etc., that is independent of the speech, but must occur because the speech reaches each of the listener's ears in segments, as a result of which some portions of the speech wave arrive in such a form that he cannot extract the cues from them. A second experiment compared the words that were correctly repeated by subjects who heard alternated speech with the words that were identified by subjects who heard only the signal for the left ear (interrupted speech) and the words that were identified by subjects who heard only the signal for the right ear (also interrupted speech, but complementary to the left‐ear signal). It was found that the performance of subjects repeating alternated speech can be adequately described as the sum of the performances of the two groups of subjects who repeated interrupted speech, at all rates of alternation, if allowance is made for the additional context available to the alternation subjects, who heard both the left‐ear signal and the right‐ear signal, one in each ear.

Perceptual Bases of Speaker Identity
View Description Hide DescriptionThis report describes all attempt to determine the number and nature of the basic ways in which voices are perceived to differ from each other by a typical listener. In the experiment, 32 listeners described their perceptions of 16 voices by means of a semantic‐differential rating form. Analysis of variance was performed to determine the contributions of speakers, listeners, and various situational parameters to the variance of ratings on each item. Factor‐analytic techniques were employed to determine the dimensionality of the speaker effect, the listener effect, and the effect of the interaction of speakers and listeners. Four factors, labeled clarity, roughness, magnitude, and animation were found to account for an average of 88% of the variance in mean ratings given speakers on each of the 49 items. Six dimensions were found to account for the common‐factor variance in constant errors associated with listeners. Five dimensions were found to account for the common factor variance of the observed interaction of speakers and listeners.

Attenuation of Guided Waves in Isotropic Viscoelastic Materials
View Description Hide DescriptionWave propagation in an infinite isotropic elastic plate with tractionfree faces is governed by the Rayleigh‐Lamb frequency equation. As is well‐known, an analysis of the equivalent viscoelastic problem shows that the frequency equation for an isotropic viscoelastic plate may be obtained from the Rayleigh‐Lamb equation by replacing the two independent elastic constants with complex functions of frequency. The propagation wavenumbers corresponding to modes that propagate unattenuated in an elastic plate are complex in a viscoelastic plate, the imaginary parts being attenuation constants. In this paper, an approximate method is developed for obtaining roots of the viscoelastic‐frequencyequation from known roots of the Rayleigh‐Lamb equation with comparatively little additional calculation. The approximate method is applicable when the attenuation per wavelength is small. Calculations have been carried out for the lowest three longitudinal and flexural modes in plates with Poisson's ratios of 0.17 and 0.35, and curves are presented that show the contributions to the over‐all attenuation owing to dissipation associated with dilatational and shear deformations separately. As an example, the variation of attenuation with frequency in a Voigt solid has been calculated and plotted for two ratios of dilatational to shear losses. The same procedure can be used to calculate the attenuation in an isotropic viscoelastic cylinder, using the Pochhammer‐Chree equation, and numerical results are given for the longitudinal modes of a cylinder.

Acoustic Impedance in a Rigid‐Walled Cylindrical Sound Channel Terminated at Both Ends with Active Transducers
View Description Hide DescriptionWhen two transducers are coupled together by a sound channel such as a water‐filled steel tube and independently driven at a common frequency, dynamic control may be exerted over the acoustic impedance presented by the channel to either of the transducers. The determination of this load in the practical case, however, presents considerable experimental difficulty. For the ideal rigid‐walled channel excited by symmetrically mounted piston sources, the load impedances may be expressed in terms either of the ratio of the source velocities or of their blocked driving forces. At frequencies below those for which radial mode propagation occurs within the channel medium, the load impedances may be also related to the standing‐wave ratio of the longitudinal mode. This latter relationship seems better suited to practical measurements. Of particular interest for underwater acousticcalibration is the use of the active‐load technique to produce plane progressive waves inside a practical water‐filled steel tube.

Excitation of a Resonant Cavity by Ducts of Small Cross‐Sectional Dimensions
View Description Hide DescriptionTwo basic configurations are considered in this paper: (1) a resonant cavity with a single‐input waveguide of vanishingly small cross‐sectional dimensions, in which the pressure in the resonator and the reflection coefficient in the waveguide are investigated. Particular emphasis is laid on the behavior around a resonant frequency of the cavity; (2) a cavity with an input and an output waveguide (a “filter”), in which the the transmission properties are investigated around resonance.

Excitation of a Circular Cylindrical Cavity by a Circular Waveguide
View Description Hide DescriptionThe problem of determining the pressure in a circular cylindrical cavity of radius b connected to a circular input waveguide of radius a is formulated in terms of an integral equation. Values of the reflection coefficient in the guide and of the pressure in the cavity are given in terms of the frequency and of the dimensionless ratio a/b. The problem of the cavity with an input and an output waveguide is also analyzed, and values of the transmission coefficient of this filter are obtained.

Absorption of Sound in Mixtures of Oxygen and Water Vapor
View Description Hide DescriptionThe attenuation of sound waves in mixtures of oxygen and water vapor contained in a tube has been measured at frequencies between 90 and 1200 cps. The tube was terminated at one end by a fixed reflector and at the other by a plane diaphragm producing the sound waves. The impedance of the diaphragm/gas‐column system was investigated near natural modes of the tube and analyzed by means of a circle diagram to yield the attenuation coefficient. The frequency of maximum absorption of sound was found to be a quadratic function of the water‐vapor concentration. The value of the vibrational specific heat for oxygen was found to be (0.0332+0.0015) R.

Transmission of Sound through a Stretched Ideal Membrane
View Description Hide DescriptionThe exact solution, in infinite series form, is given to the problem of the transmission of an axially symmetric sound wave through an ideal stretched membrane in a rigid circular tube. Certain simple exact results are deduced, and the exact equations for the resonance and antiresonance frequencies are considered. It is shown that, for an incident plane wave, above a certain frequency there are no exact resonances or antiresonances. Below this frequency, the resonances and antiresonances differ from those predicted by approximate methods, and, contrary to the simple results, their frequencies depend on the density of the medium surrounding the membrane. The same analysis, which involves the inversion of an infinite matrix, applies also to the parallel‐plate guide and to transmission through a thin elastic plate. Numerical results are given for a particular case.

Validity of Thin‐Plate Theory in Dynamic Viscoelasticity
View Description Hide DescriptionResonance damping for a vibrating plate is investigated both according to the exact equations of dynamical viscoelasticity and the classical thin‐plate equations derived in mechanics of materials. The plate is assumed as isotropic and homogeneous and no shear‐ or rotatory‐inertia corrections have been included in the thin‐plate approximations. Two types of materials are investigated that correspond to real and complex values of the bulk modulus. For each case, the complex shear modulus is μ(1+ig) and values of g up to 0.10 were used in the calculations. The two theories are in excellent agreement in a range of wavelengths as low as about ten times the thickness. It is found that thin‐plate theory evaluates the damping more accurately than it does the static rigidity.

Method of Perturbation Applied to the Vibration Problem of a Circular Membrane of Varying Density
View Description Hide DescriptionIn this paper, a treatment for determining the eigenfunctions and eigenvalues for the vibration of a circular membrane with fixed boundary and density varying as exp[− εγ^{2}] is made by the method of perturbation, and the expressions for perturbed eigenvalues and eigenfunctions are obtained up to second‐order terms. The first‐order perturbed term of the eigenvalue is effected by numerical integration.

Steady‐State Behavior of the Dynamic Absorber—Addendum
View Description Hide DescriptionDiscussed theoretically is the manner in which the transmissibility of a vibrating item may be reduced by the attachment of a dynamic absorber, or by the attachment of three absorbers of staggered natural frequencies. It is demonstrated that the dynamic absorbers can provide significant isolation at frequencies of concern; yet the presence of the “compensating” resonant peaks normally associated with the use of a dynamic absorber may be avoided by employing small absorber damping and by damping heavily the resilient elements that support the vibrating item.

Analysis of Echoes from a Hollow Metallic Sphere in Water
View Description Hide DescriptionThe nature of sonarechoes from a hollow metallic sphere in water is investigated theoretically in relation to the thickness of the shell and the contents of the hollow region. Steady‐state solutions are presented for iron and aluminum spheres filled with water and for iron spheres with a vacuum inside. Calculated results for pulsed sound suggest that part of the echo originates from a type of flexural wave moving around the shell, which changes into a surface wave as the shell thickness increases. Significant differences in echo structure occur when the interior region is changed from water to a vacuum.

Underwater Noise Produced by Bullet Entry
View Description Hide DescriptionInvestigations of the underwater sound produced by shooting bullets into water indicated that the principal mechanism of sound production might be the oscillatory collapse of an air bubble near the surface. High‐speed motion pictures taken of a .38‐caliber‐pistol bullet's entry into a tank confirmed this idea. The main features of the acoustic signature can be explained and correlated with the motion pictures. On the order of 1% of the available kinetic energy was converted to acoustic energy. Most of the energy was transferred to a single bubble during the first 2 ft of travel in the water.

Propagation of Explosive Sound in a Layered Liquid
View Description Hide DescriptionThe transient pressure produced by explosive sources in a two‐layered liquid half space has been evaluated in terms of definite integrals, using Cagniard's method of inversion of the Laplace transforms involved.

Model of a Rough Boundary as a Backscatterer of Wave Radiation
View Description Hide DescriptionIt is hypothesized that the backscatter of wave radiation from a rough boundary is the result of reflections from a large number of “facets” having random distributions of size and slope. This hypothesis developed into a mathematical model from which curves are computed for a number of values of mean size and rms slope. An explanation is offered for the lack of frequency dependence in reported backscattermeasurements. The model is also fitted to some previously reported measurements of the backscatter of sound from the deep‐ocean bottom.

Self‐ and Mutual‐Radiation Impedance and Beam Patterns for Flexural Disks in a Rigid Plane
View Description Hide DescriptionExpressions are given for the nearfield pressure on a flexural disk in an infinite, rigid, plane baffle. This pressure is then integrated over the disk to obtain finite series expressions for self‐radiation impedance, which is referred to average surface velocity, as are all impedances in this article. Expressions are derived for the far field directionality factor of a flexural disk. The Bouwkamp integral method is used to obtain infinite series expressions for mutual radiation impedance between flexural disks by integrating the farfield directionality factor of two identical flexural disks, vibrating in phase, over a set of complex angles. It is shown that as the relative radius (ka) of the disk approaches zero and the element spacing becomes large as compared to the radius of the disk, the equations for mutual‐radiation resistance and reactance and self‐radiation resistance for flexural disks are the same as for circular pistons. However, as ka → 0, the self‐radiation reactance is larger for disks than pistons, being 0.849 ka for pistons, 1.060 ka for supported‐edge disks, and 1.204 ka for clamped‐edge disks.

Effects of Extraneous Tasks on Auditory Fatigue
View Description Hide DescriptionNine subjects with normal hearing were exposed for 3 min to 50‐ and 100‐dB‐SPL, 4‐kc/sec tones during the performance of a task. Three tasks were used: long division, light tracking, and reverie. Each subject received 3 replications of every task/level‐of‐exposure combination. Performance of task during exposure was found to have no significant effect on the magnitude of TTS. Repeat‐reliability measures showed the subjects to be reasonably consistant in their responses.

Temporary Hearing Losses following Exposure to Pronounced Single‐Frequency Components in Broad‐Band Noise
View Description Hide DescriptionTemporary hearing losses were observed for 20‐min exposures to pure‐tone frequencies (500, 1000, 2000, 4000 cps), which were independently mixed in various strengths with a broad‐band noise to create different exposure conditions involving a strong tonal component in a noise field. The over‐all (combined pure‐tone/noise) SPL for all conditions was fixed at 105 dB and the resultant losses were compared to those caused by equivalent exposures to just the broad band noise (referred to as the continuous‐spectrum‐noise condition). The pure‐tone/noise combinations produced greater threshold losses than the continuous‐spectrum noise primarily when the pure tone was of low frequency (below 2000 cps) and contained most of the available energy in the sound field. These more prominent pure‐tone exposures exceeded the criteria set forth by the U. S. Air Force 160‐3 Noise Regulation for identifying those conditions of strong pure‐tone energy in noise that are believed more hazardous to hearing. However, other pronounced pure‐tone conditions in this study also met these criteria even though they gave no evidence for increased hazard to hearing when their postexposure losses were compared with those of the continuous‐spectrum noise. The findings were discussed with regard to the action of the acoustic reflex that has been proposed to account for the differences in the ability of pure tones and noise to cause hearing loss.

Measurement of Auditory Density
View Description Hide DescriptionAuditory density refers to the apparent compactness, concentration, or hardness of a sound. An attempt was made to determine directly the form of the density surface by the method of magnitude estimation with both pure tones and bands of noise (quarter‐octave). For both tones and noises, density grows as a power function of sound pressure. The values of the exponents decrease with frequency, and the two families of power functions converge toward a common point near about 150 dB SPL. The equal‐density contours turn out to be straight lines in a plot of SPL vs log frequency. The general slope of the equal‐density contours for pure tones was verified by a direct‐matching procedure.

Covariance of Noise in Attenuating Media
View Description Hide DescriptionIn this paper, the covariance function of noise at two points in a sound‐carrying medium is derived for an arbitrary distribution of noisegenerators when the noise suffers attenuation and spreading loss. The analysis is then applied to the cases of a volume and a plane distribution of omnidirectional sources. It is shown that the effect of attenuation on covariance is generally negligible in the volume case but is ordinarily important in the plane case. For the plane distribution, an asymptotic expansion is derived that holds for small values of an attenuation parameter. This expansion is then used to study the effect of attenuation on covariance as a function of receiver separation, receiver‐line orientation, and time delay.