Volume 22, Issue 4, July 1950
Index of content:
22(1950); http://dx.doi.org/10.1121/1.1906618View Description Hide Description
Phase comparison at high ultrasonic frequencies is found to be particularly suited to velocity of propagation measurements (and hence measurement of elastic constants) for small solid specimens such as single crystals. Both longitudinal and transverse waves may be used, with accurate results possible even when the specimen loss becomes high.
A pulse technique is used which allows exact determination of the number of wave‐lengths in the acoustic path, and a direct measurement of phase shift at reflecting interfaces. The use of λ/4 plastic coupling seals to minimize errors is explained; necessary theoretical analysis is included.
Illustrative data for fused silica and single crystal germanium are given.
22(1950); http://dx.doi.org/10.1121/1.1906619View Description Hide Description
An approximate theoretical treatment of the output from the simplest type of ultrasonic generator in the form of a laminated bar is first developed. The dumbbell generator is shown to be very similar to the simple bar type and the previously obtained relations are then employed for determining the relative output in terms of its face and neck dimensions.
22(1950); http://dx.doi.org/10.1121/1.1906620View Description Hide Description
The acoustic absorption of sand and soil with varying amounts of water (dry to saturated) has been investigated the 10‐ to 100‐kc range. Two methods were used. In one, the source and a number of equally spaced probe microphones were imbedded in the medium. In the other method the source and receiver were in air and the transmission loss was obtained for samples of varying thickness. The unsaturated media were found to have attenuation coefficients ranging from 2 db/cm for nodulous, loose soils to greater than 25 db/cm for finely divided soils and sand, the value for any particular sample depending in an important way on its flow resistance. In water‐saturated media the attenuation was found to depend markedly on the amount of gas present in the mixture. Air‐free mixtures prepared in an evacuated chamber had attenuation coefficients which were too small to be measurable—probably less than 2 db/cm. On the other hand, for mixtures prepared in the presence of air the attenuation coefficient immediately after mixing were extremely large. For one sample it ranged from 26 db/cm at 10 kc to 64 db/cm at 30 kc. When a source and microphone, separated by a distance of one centimeter, were immersed in such a freshly stirred mixture and then left undisturbed, the intensity level increased in time—as much as 50 db at 30 kc over a period of 100 hr.
22(1950); http://dx.doi.org/10.1121/1.1906621View Description Hide Description
The scattering of sound from a spherical fluid obstacle of size comparable to a wave‐length is considered, neglecting dissipation. Calculations of the acoustic pressure and the total energy in the scatteredwave are presented graphically; soundvelocities and densities of the sphere lie between 0.5 and 2.0 times that of the external medium. The limiting cases of Rayleigh scattering and scattering from a fixed rigid sphere are also shown for comparison. In the region where the diameter of the sphere is comparable to a wave‐length, the scattering is a complicated function of frequency, showing in some cases large maxima and minima. The amplitude of the scatteredwave in the backward direction from a fluid sphere a few wave‐lengths in diameter exceeds twice that from a rigid sphere of the same size for the case of the soundvelocity 0.8 and density equal to that of the surrounding medium.
22(1950); http://dx.doi.org/10.1121/1.1906622View Description Hide Description
A method for the calculation of viscous losses in acoustic wave guides and cavities is described, similar to that used by Carson, Mead, and Schelkunoff for electromagnetic waves. The loss in a plane wave in a round tube is discussed and the results agree with those previously obtained by others. The attenuation for two higher modes in hard wall guides is computed, as well as the decay constants for a cylindrical cavity for longitudinal and pure radial modes.
22(1950); http://dx.doi.org/10.1121/1.1906623View Description Hide Description
The flexural vibrations of a ring in its plane are studied by taking shear effect into account, and it is proposed to compute the deviation from the classical frequency formula, which is only valid for very small ratios of the thickness to the diameter. In a second chapter the possibility of extension is included, and in this procedure the frequencies of extensional vibrations are obtained as a by‐product.
It is seen that the shear effect plays a predominant role in the deviation from the classical formula, and is more important than curvature effects and the rotational inertia. Extensibility influence is found to be very small.
An interesting comparison can be made with the similar influence of secondary factors on the flexural vibration's frequencies of a straight beam.
22(1950); http://dx.doi.org/10.1121/1.1906624View Description Hide Description
Direct measurements of envelope delay have been obtained for a series of commercial and experimental loudspeakers. The modulation phase shift method of Nyquist and Brand has been adapted for direct indication of envelope delay by utilizing an electronic phase meter. A continuous record of the delay is obtained by an automatic level recorder suitably connected to the phase meter. Pressure‐amplitude curves measured under the same conditions provide data for a preliminary attempt to interpret the correlation between delay response and pressure response.
22(1950); http://dx.doi.org/10.1121/1.1906625View Description Hide Description
A heterodyne filter for the speech frequency range is described. It can be used as a band pass filter of continuously variable low and high frequency cut‐offs adaptable for any desired transmission band in the frequency range 40–4000 c.p.s. and with more than 60‐db attenuation and with cut‐offs of a maximal slope of 1 db per c.p.s.
Band width and low frequency cut off are varied by two separate controls. Band elimination filtering can also be obtained with continuously variable mid‐frequency but with band width restricted to either 300 or 800 c.p.s.
The variable filtering is performed by single sideband transmission with successive filtering in two modulation units and demodulation in a third unit. Distortion products from impurities in transmission and modulation are kept at a −60‐db level and random noise at a −85‐db level, relative to a single test tone.
The filter was designed for articulation tests and general speech research purposes and also as a waveanalyzer of continuously variable band width, 45–4000 c.p.s. in the frequency range 40–2000 c.p.s. The maximum permissible speed in analysis of sustained sounds, governed by the inverse square of band width law, is discussed with regard to the quality of the analyzing band pass filter.
22(1950); http://dx.doi.org/10.1121/1.1906626View Description Hide Description
A technique is described for the dynamic calibration of accelerometers or other vibration‐measuring instruments. For each calibration frequency the accelerometer is attached to a free‐free bar driven electromagnetically at a known sinusoidal displacement. Displacement amplitudes are measured by means of strain gauges which are calibrated optically. The accelerometer output can be conveniently compared in magnitude and phase with the displacement as indicated by the strain‐gauge voltage. Instruments have been calibrated at frequencies between 400 and 2000 c.p.s. The error in determining phase angle was about ±2 degrees. The error in the optical measurement of displacement was about ±0.1 × 10−3 cm. The percent error, however, could be made fairly small by calibrating the strain gauges at large displacements of the bar. For instance, the error in calibrating a 200‐g accelerometer at 1000 c.p.s. was about ±2 percent.
22(1950); http://dx.doi.org/10.1121/1.1906627View Description Hide Description
One hundred and ten monosyllabic words, 10 for each of the 11 common American vowels, were spoken in isolation by each of 10 subjects. Most of the differences between the mean relative intensities of the vowels were found to be statistically significant. Among the words for a given vowel the intensity of that vowel was found to differ significantly in most instances, variation tentatively attributed to consonantal environment.
22(1950); http://dx.doi.org/10.1121/1.1906628View Description Hide Description
Measurements are made of the relative effectiveness of the round window as a route of entrance of sounds to the cochlea. When the ear is normal this route is of no importance, but when the middle ear apparatus is absent and its advantage is lost to the oval window route a sound will have nearly equal access by both windows. When the inner ear is reached by both pathways at once the cochlear potentials represent the vector sum of what would result from the two separate waves. As phase and intensity relations between the two pathways are altered the potentials pass through maximum and minimum values, which for equal intensities of the two waves vary from a 6 db gain to a complete loss.
Evidence is produced to show that each pathway of stimulation excites the same sensory cells and in the same intensity pattern.
Over the major portion of the frequency range a minimum of response results when the waves are in phase as they enter the oval and round windows. Departures from this relation, which are most prominent at the highest frequencies, are explained as the result of reactance differences between the two windows.
These results apply to the explanation of forms of deafness caused by interruption of the ossicular chain, and also to the explanation of the treatment of otosclerosis by means of the fenestration operation.
22(1950); http://dx.doi.org/10.1121/1.1906629View Description Hide Description
The report describes measurements of the frequency limits of binaural beats and outlines a theory of binaural beats, based on synchronous discharges in the two auditory nerves. Two sinusoids of frequencies f 1 (fixed) and f 2 (variable) were led separately to the two ears; and the difference that marked the disappearance of the fluctuating loudness or roughness that is characteristic of binaural beats was determined. Δf was maximal (approximately 35 c.p.s.) for frequencies in the neighborhood of 400 c.p.s. Binaural beats were heard above 1000 c.p.s., but careful attention was required and Δf was small. The shape of the curve relating Δf to f 1 provides an explanation for the fact that determinations of the upper frequency limit of binaural beats have not been in agreement; the upper frequency limit depends markedly on Δf. The theory, given to account for the fact that Δf is smaller both at low and at high frequencies than it is near 400 c.p.s., combines elements of the Hill‐Rashevsky theory of the excitation of neurons with elements of Wever's volley theory. At low frequencies neurons can discharge in some degree of synchrony with the stimulus wave form, yet fail to coincide within the time interval necessary for synaptic summation. At high frequencies the neurons must take turns discharging, and relatively few can participate in any given volley. At intermediate frequencies, however, each neuron participates in many volleys and the neurons participating in each volley fire almost simultaneously. The result is that at intermediate frequencies synchrony is relatively precise in each afferent pathway and, when the two afferent streams join in a common neural center, beats appear.
22(1950); http://dx.doi.org/10.1121/1.1906630View Description Hide Description
A phonographically recorded test of the ability to hear pure tones was given to 3666 people at the San Diego County Fair in the summer of 1948. Absolute thresholds were determined at five frequencies and masked thresholds at two frequencies. The results were analyzed according to the age, sex, musical training, and past noise environment of the listeners, and according to their statements as to whether or not they had difficulty in hearing. As expected, auditory sensitivity was found to decline with age; women were found to be more sensitive than men for the higher frequencies; and men who worked in noise showed greater than normal losses at high frequencies. The data indicated, in addition, that musically trained men and women possess greater hearing sensitivity than do men and women without musical training, and that the 20–29 year old males tested at San Diego in 1948 appeared to have a greater loss at the high frequencies than did most males of the same age group who were tested in surveys before the war.
22(1950); http://dx.doi.org/10.1121/1.1906631View Description Hide Description
Absolute thresholds and thresholds of pure tones masked by thermal noise were obtained for 3666 persons attending the 1948 San Diego County Fair. The masked hearing losses as well as the absolute hearing losses (1) increased with an individual's age, (2) were positively correlated with his awareness of hearing difficulty, and (3) were inversely related to his musical training. Masked losses, however, were less dependent upon one's sex and upon one's past noise environment than were absolute losses.
When groups of individuals were equated as to their absolute losses on 880 c.p.s. and 3520 c.p.s. (the frequencies which were masked), that part of each group that had the lower absolute thresholds on 440 c.p.s. also had the lower masked thresholds.
The signal to noise ratio (noise expressed as spectrum level) at masked threshold for those with no absolute loss was 20 db at 880 c.p.s. and 21 db at 3520 c.p.s. This is equivalent to saying that the noise bands which were effective in masking these frequencies were 100 and 128 cycles wide, respectively.
22(1950); http://dx.doi.org/10.1121/1.1906632View Description Hide Description
The results of experiments on masking of tones by narrow bands of noise at 200, 800, and 3200 c.p.s. are presented. These are interpreted in terms of the critical band hypothesis. Assuming a simple resonant characteristic for the aural frequency selectivity the shape of the masking vs. frequency function is computed and gives a reasonably good fit in the narrow regions studied. The critical band widths are determined at frequencies 200, 800, and 3200 c.p.s., and the corresponding values of Q of the equivalent tuned circuit are given. Data are presented for masking of tones slightly removed from the masking noise band, and are shown in fair agreement with the computed values. The selectivity curves are shown to be consistent with available data on differential sensitivity to frequency.
22(1950); http://dx.doi.org/10.1121/1.1906633View Description Hide Description
When a pure tone and a masking noise are fed into the same ear simultaneously, the pitch of the pure tone is raised. The effect is more pronounced at low loudness levels and is progressively greater as the frequency of the tone rises. Apparently the phenomenon is not present if the pure tone is 20 db or more above its masked threshold. Nineteen graduate music students served as subjects in the experiment. Control measurements were made to estimate the effects of diplacusis and of the difference in intensity between the masked and the unmasked pure tone when they were matched in loudness.
22(1950); http://dx.doi.org/10.1121/1.1906634View Description Hide Description
The magnetostrictiveproperties of Hiperco, a magnetic alloy of iron,cobalt, and chromium, were investigated to determine its suitability as a material for electromechanical transducers. The relations between the magnetic and magnetostrictivecharacteristics of a material and its performance as a transducer element are reviewed and criteria for its evaluation established. The properties of Hiperco, measured after a suitable heat treatment had been determined, are discussed in relation to these criteria and a comparison made with nickel and Permendur. For operation at magnetic remanence Hiperco is shown to be as good as Permendur for hydrophone applications, but inferior to nickel and Permendur when used as an underwater projector at high power levels. When polarized above remanence maximum value of 0.17 for the electromechanical coupling coefficient is realized and operation as a projector is improved. Performance data are given for a small underwater transducer constructed of Hiperco operating at remanence.
- LETTERS TO THE EDITOR
22(1950); http://dx.doi.org/10.1121/1.1906635View Description Hide Description
Studies of the response of an auditorium to a short tone burst have indicated that the character of the first 20 or 30 db of sound decay, the “short‐term” response, is closely related to the subjective “hearing quality” of the room. Recently Haas has investigated the effect of a single echo on the subjective hearing of speech. On the basis of some of Haas' results and other information on the hearing of speech in rooms, a tentative criterion for the short‐term response has been formulated as curves of amplitude vs. time delay of each reflection relative to the direct sound. The criterion correctly rank‐orders subjective quality observations reported by others, and agrees quantitatively with some subjective judgments reported here.
Report on Six Lectures by Richard H. Bolt to the Physical Society of London on the Acoustics of Rooms22(1950); http://dx.doi.org/10.1121/1.1906636View Description Hide Description
22(1950); http://dx.doi.org/10.1121/1.1906637View Description Hide Description