Volume 23, Issue 3, May 1951
Index of content:
23(1951); http://dx.doi.org/10.1121/1.1906756View Description Hide Description
An equivalent‐tone method for calculating the loudness of sounds is described. With this method the spectrum of the sound is divided into frequency bands which are treated as pure tones in determining their loudness. The individual values of loudness are added to obtain the total loudness of the sound. Calculations for bands of white noise and for complex tones are compared with subjectively obtained data of Pollack and Fletcher and Munson. The agreement between calculated and experimental values is good. It is felt that improvement of the method must await further psychoacoustic data.
23(1951); http://dx.doi.org/10.1121/1.1906757View Description Hide Description
In the course of previous investigations, Dix, Hallpike, and Hood have concluded that the phenomenon of loudness recruitment is attributable to disordered function resulting from injury or disease of the hair cells of Corti's organ.
In the present paper, an account is given of an experimental study of certain loudness changes caused by adaptation of the hair cell responses to pure tone stimulation.
Two outstanding characteristics of the adapted state are defined: “on‐effect normality” and “relapse.” Both resemble very closely those described by Matthews in the case of adaptation occurring the muscle stretch receptors of the frog.
Further studies in a number of human subjects show that both of these characteristics are exhibited regularly by the unadapted hair cell responses of individuals suffering from degenerative changes of the hair cells.
These studies appear to establish a very close similarity between “on‐effect normality” occurring in the case of the adapted normal sense organ and the phenomenon of loudness recruitment, as demonstrated by the alternate binauralloudness balance procedure of Fowler in the case of the unadapted but diseased sense organ.
The experimental findings are considered to favor the view that loudness recruitment is one element of a complex disturbance of cochlear function due to disease or injury of specific elements of the cochlear sensory apparatus, namely the hair cells of Corti's organ.
23(1951); http://dx.doi.org/10.1121/1.1906758View Description Hide Description
The method of best beats has been employed to estimate the intensities of aural harmonics and of combination tones. It has been generally assumed that the listener hears best beats when the exploring tone produces in the cochlea a disturbance that is equal in magnitude to that of the aural harmonic or of the combination tone being measured. However, as the experiments to be reported will show, when the tone to be measured is near the absolute threshold or is partially masked, the most prominent beats will be heard when the intensity of the exploring tone exceeds that of the unknown tone. Consequently, since aural harmonics and combination tones are partially masked, their intensities will be overestimated when the method of best beats is used. The present paper concerns the magnitude of this error of overestimation.
A procedure is presented by which a better estimate of the intensity of an aural harmonic or of a combination tone may be made. When the tone being measured by the method of best beats is well above threshold, the listener hears beats over a wide range of intensities of the exploring tone, and the error of over‐estimation is small. However, when the tone being measured is near threshold, there is a small range of audible beats, and the error of overestimation is of considerable magnitude. It is therefore possible to correct for this error by determining the range of intensities of the exploring tone over which beats are audible as well as the intensity of the exploring tone required for best beats. Application of this procedure to the measurement of a second aural harmonic is illustrated.
An explanation is given of the fact that the intensity of a tone near its absolute threshold will be overestimated by use of the method of best beats. This explanation is formulated in terms of the relation of the minimum and the maximum of the envelope of the beating complex to the listener's threshold for the tones that beat.
23(1951); http://dx.doi.org/10.1121/1.1906759View Description Hide Description
A discussion is made of two lines of evidence that bear upon the problem of the relative importance of place and frequency principles in auditory theory. The first line of evidence has to do with the relation between pitchdiscrimination and the degree of specificity of action in the cochlea and brings us to an evaluation of the principle of maximum stimulation. A consideration of loudnessdiscrimination data discloses a serious limitation on our ability to appreciate the point of maximum in the cochlear activity and, in general, to make use of a spatial pattern. We must conclude that for the low tones whose patterns are broad the place principle contributes little or nothing to pitchdiscrimination, and the frequency principle has to serve this function alone.
A study of the evolution of the ear brings out still more clearly the relative roles of place and frequency principles. Hearing in some degree is present throughout the vertebrate series, and even the lowest forms have frequency ranges of several octaves and exhibit a fair degree of tonal discrimination. Yet, these lower forms possess an ear of exceeding simplicity, an otic sac in which there is little evidence of mechanical differentiation. It follows that these primitive ears must operate as frequency receptors, with ranges and discriminative abilities that depend simply upon the patterns of impulses conveyed by the auditory nerve. It appears, therefore, that in the history of the ear the frequency principle came first as the basis of tonal reception, and only sometime later in the course of evolutionary development was it joined by the place principle.
23(1951); http://dx.doi.org/10.1121/1.1906760View Description Hide Description
One of the central problems in auditory theory is to reconcile (a) the acute perception of slight changes in pitch displayed by the human listener with (b) the broad tuning of his cochlear analyzing mechanism. This paper attempts to describe and to relate a number of theoreticalsolutions to that problem. The hypotheses involve mechanisms, both mechanical and neural, for sharpening the analysis inherent in the cochlear transformation from frequency of stimulation to locus of vibration. These mechanisms operate in the domain of place—they are place theories that supplement the classical place theory. In a future paper, we plan to describe and discuss other sharpening mechanisms that operate in the domain of time.
23(1951); http://dx.doi.org/10.1121/1.1906761View Description Hide Description
An artificial mastoid for testing bone conduction receivers is described. It consists essentially of a stiff metal bar which has a fundamental resonant frequency above the measurement range, with a strain‐sensitive translating element fastened to the under side and a compliant pad on top.
The translating element is a slab of activated ceramic, which is essentially invariant to humidity and temperature changes. The compliant pad generally used is Koroseal No. 74, which simulates the flesh over the mastoid prominence reasonably well. Koroseal No. 15 simulates flesh better but is less rugged. Calibration methods are discussed.
23(1951); http://dx.doi.org/10.1121/1.1906762View Description Hide Description
Speech articulation and quality tests were made on an all‐pass system capable of advancing or delaying one frequency band relative to the rest of the spectrum. Parameters in the investigation were (a) width of the advanced or delayed band, (b) amount of advance or delay, and (c) position in the frequency spectrum of the advanced or delayed band. Data were taken at signal‐to‐noise ratios of 30 db and 0 db. The results indicate that maximum impairment of speech intelligibility and quality occurs when the delays and advances are of the order of , and when the band that is advanced or delayed is near the center of the speech spectrum and has an articulation index equal to 0.5. These findings are related to data from statistical studies of the timing of speech energy bursts.
23(1951); http://dx.doi.org/10.1121/1.1906763View Description Hide Description
The fundamental wavelength of a spherical resonator with a circular aperture is calculated. The result, , where a denotes the aperture and R the sphere radius, is the correct form of an expansion as far as terms of relative order (a/R)2, inclusive. A procedure for approximating to the wavelength of a resonator with arbitrary shape is also described.
23(1951); http://dx.doi.org/10.1121/1.1906764View Description Hide Description
A general expression is derived for the force owing to radiation pressure acting on an object of any shape and having an arbitrary normal boundary impedance. It is shown that boundary layer losses may lead to forces that are several orders of magnitude greater than the forces owing to classical radiation pressure. Steady forces arising from an asymmetric wave form are compared with the other forces. A sound wave, consisting of equal parts of fundamental and second harmonic components, can cause forces ten or more orders of magnitude greater than the forces owing to radiation pressure to be exerted on small particles.
23(1951); http://dx.doi.org/10.1121/1.1906765View Description Hide Description
A method for treating the scattering and absorption of sound in the presence of non‐uniform boundary conditions is applied to the case of an infinite strip of material (of given width) placed on an infinite, otherwise hard, wall. The strip is assumed to be characterized by a normal acoustic impedance that may possess a resistive component.
The wave equation is reformulated as an integral equation over the strip. A variational expression is found for the amplitude of soundscattered by the strip in any direction. The total cross section for scattering plus absorption is determined from the amplitude of specular reflection according to a well‐known cross‐section theorem. Comparison with Pellam's results for absorption by such a strip shows that the methods used have an accuracy better than 2 percent over the entire frequency range.
23(1951); http://dx.doi.org/10.1121/1.1906766View Description Hide Description
Sound scattering from a sphere of arbitrary size is treated theoretically when the acoustic properties of the sphere are near those of the surrounding medium. Closed form analytic expressions are found for the reflectivity and total cross section. These expressions become exact only in the limit as the ratio of the densities and ratio of the speeds approach unity, but it is shown that for ratios as large as 5/4 and probably as large as 3/2 the approximate reflectivity and total cross section compare favorably with results of exact calculation. Calculations of reflectivity and total cross section, which would require weeks if made from the exact solution of the wave equation, may be completed in a few hours using the approximate closed form expressions.
23(1951); http://dx.doi.org/10.1121/1.1906767View Description Hide Description
The reflection of a spherical sound wave from a wall with the boundary conditions expressed in terms of a normal impedance independent of the angle of incidence is treated. It is shown that the integral for the reflected wave can be evaluated exactly in closed form under certain conditions. The solution given for an arbitrary normal impedance involves a slight approximation of the integral. The reflected wave is brought into a form such that it can be considered originating from an “image source” having a certain amplitude and phase. Graphs for determining this amplitude and phase are given in terms of a “numerical distance,” which depends on the normal impedance and the position of the field point. Pressure distributions around point sources for different wall impedances are shown. The limitations in simulating plane wave conditions at a boundary and the corresponding effect on free field methods of measuringacoustic impedance are discussed.
23(1951); http://dx.doi.org/10.1121/1.1906768View Description Hide Description
The nonspecular reflection of plane waves of sound from certain surfaces composed of absorbent bosses (semicylinders or hemispheres of arbitrary impedance) on an infinite plane of ∞ or 0 impedance is considered. Exact solutions are obtained for the problem of the single boss and then extended, subject to the single‐scattering hypothesis, to obtain far field solutions for certain planar distributions of bosses of radii small compared with the wavelength. The results are compared with those obtained previously for non‐absorbent bosses, and it is shown that the effects of the finite impedance are most pronounced in the simple source terms of the scattered components and may lead to either a decrease or an increase in the radiation reflected at the specular angle. Another effect of the finite impedance (for the small finite distributions) is to shift the critical value of the angle of incidence for which the reflection at the specular angle consists only of the specular component—below this value the reflection at the specular angle being a minimum and above it a maximum. For the infinite uniform random distributions it is found that the effect of the bosses is essentially but to change the impedance of the plane—these effective impedances being functions of the angle of incidence and the parameters and distribution of the bosses. The effect of the finite impedance of the bosses is most pronounced for these distributions yielding terms much lager than those previously retained for the nonabsorbent bosses. The results for the analogous distributions of cylinders and spheres are also given.
23(1951); http://dx.doi.org/10.1121/1.1906769View Description Hide Description
The measured transmittivity of a steel plate in water is presented as a function of the angle of incidence and the product of frequency and plate thickness over wide ranges.
The normal velocity of the plate surface can attain an amplitude necessary for good transmission only by constructive interference among internal reflections. It is shown that the ideal conditions can be met in a plate of finite width in only a few cases. In general, the conditions for a transmission maximum are the conditions for the existence of appropriate types of stable traveling waves in a plate of infinite extent; these conditions, however, are modified by edge effects. An apparent effect of this modification is to produce changes in the divergence of the transmitted beam and hence in the observed transmittivity.
23(1951); http://dx.doi.org/10.1121/1.1906772View Description Hide Description
It is assumed that acoustic background noise is caused by a distribution of random “white” noise sources whose physical mechanism is unspecified. A law expressing the amplitude‐distance attenuation characteristic of the medium is also assumed. Several distributions of noise sources are considered: uniform volume distributions, uniform surface dipole distributions, and two mixed cases. The noise drop‐off with frequency at a point below the surface is found for each case. For an infinite volume of noise sources, this drop‐off is 6 db/octave at all frequencies. It is shown how this simple model can be generalized to other attenuation laws and other spatial and amplitude distributions of noise sources.
23(1951); http://dx.doi.org/10.1121/1.1906773View Description Hide Description
To obtain information the characteristics of relaxation phenomena in liquids, measurements of ultrasonic absorption were performed in seven binary mixtures having nitrobenzene as one component.
The absorption coefficients in some systems (benzene, chloroform) confirm the results obtained previously in mixtures formed by an unassociated, very absorbing liquid and another component with much smaller absorption coefficient. In this case, the absorption coefficient decreases quickly when small quantities of the second liquid are added to the first one. The explanation of this behavior seems to be found in the fact that an increase of the efficiency of collisions between the molecules occurs. When different molecules are added to the high absorbing liquid, they decrease the time necessary to establish equilibrium among the internal degrees of freedom of the high absorbing molecules.
In other mixtures the two components have similar absorption coefficients. These systems are of great interest, because the characteristic behavior of the ultrasonic absorption in the mixture then depends on the nature of both pure liquids. If the two liquids are unassociated and their molecules do not interact strongly between themselves, the curve of the absorption coefficientversus concentration presents a very clear minimum. If, instead, there are strong interactions, the shape of the curve is altered. This would indicate the presence of an additional type of energy loss.
In the systems of nitrobenzene and an alcohol, the absorption coefficient, as a function of the mole fraction, has a maximum at an intermediate concentration, just as happens in some mixtures of water and alcohols.
Physical Factors Involved in Ultrasonically Induced Changes in Living Systems: II. Amplitude Duration Relations and the Effect of Hydrostatic Pressure for Nerve Tissue23(1951); http://dx.doi.org/10.1121/1.1906774View Description Hide Description
The results of experiments with frogs under a hydrostaticpressure demonstrate that cavitation is not an important factor in the mechanism of production of paralysis of the hind legs of frog by ultrasonic (frequency one megacycle) irradiation over the lumbar enlargement region of the spinal cord. Experimental results indicate that a linear relation exists between the reciprocal of the minimum exposure time for paralysis and the acoustic amplitude. This result is readily described in terms of a one factor rate process. On the basis of this experimentally determined relation, it is shown that time rate of change of temperature cannot be correlated with the observations. It is concluded on the basis of a theoretical calculation that absorption of ultrasound at interfaces in the spinal cord does not result in minute hot regions.
Further work on summation of subparalytic doses, spaced apart at various time intervals, indicates that the recovery process following exposure to a subparalytic dose of ultrasonic radiation may not be a monotonic function of time.
- LETTERS TO THE EDITOR
23(1951); http://dx.doi.org/10.1121/1.1906775View Description Hide Description