Volume 20, Issue 2, March 1948
Index of content:
- PROGRAM OF THE THIRTY‐FOURTH MEETING OF THE ACOUSTICAL SOCIETY OF AMERICA
20(1948); http://dx.doi.org/10.1121/1.1916867View Description Hide Description
By the most comfortable listening level for a pure tone is meant the level which the listener chooses as most comfortable for listening when he listens to a pure tone of which he can change the level at will. Ranges and mean values of most comfortable listening levels are given for pure tones in the range 128 to 8192 c.p.s. (at octave intervals) for listeners with normal hearing and listeners previously exposed to gunfire and airplanenoise, with resultant marked hearing loss in some cases. Tentative most comfortable listening level contours are suggested. The use of a most comfortable level contour for specification of a hearing aid amplifier is discussed and an example given of the effectiveness of such an amplifier as compared to one specified by a most comfortable loudness level contour.
20(1948); http://dx.doi.org/10.1121/1.1916906View Description Hide Description
Bone conduction thresholds were taken at frequencies from 1500 to 12,000 c.p.s., using the induction‐type electrodynamic bone conductor and the air and bone conduction test assembly described in earlier papers. Stray air radiation from the bone conductor is an important factor at the frequencies considered; therefore, the bone conductor was enclosed by a sponge rubber jacket to minimize the air radiation, and the ears were covered by the test assembly boxes lined with absorbing material in order to protect them from the residual radiation. The present investigation is an extension of earlier work done at frequencies at which stray air radiation effects are negligible, and the results obtained are combined with those of the earlier study to yield a tentative bone conduction threshold from 80 to 12000 c.p.s.
20(1948); http://dx.doi.org/10.1121/1.1916915View Description Hide Description
Equations showing the relation between the volume, V, of an auditorium and its optimum reverberation time, T 0, were published independently by MacNair and Liftschitz. MacNair's equation was derived from the following basic assumptions: , (1) , (2) where ρ = instantaneous value of the energy density of sound during decay. T 1 = time of decay to minimum audibility. E = rate of supply of energy during the steady state condition.
This paper shows that an equation, relating T 0 and V, can be deduced which more nearly fits all the empirical facts by basing the derivation on the following assumptions: , (1) , (2) where G = a constant, n = a constant which should be less than unity.
The resulting equation, for , is . This is in agreement with the empirical data and with the equation derived by Maxfield and Albersheim from acoustic “liveness” considerations.
The psychologists and physiologists studying the mechanism of hearing should be interested in these results, as they have a bearing on the manner in which the ear integrates the sound being perceived.
20(1948); http://dx.doi.org/10.1121/1.1916916View Description Hide Description
Some experiments have been conducted to determine the effect of intense high frequency airborne sound on mice and a variety of insects. The sound source was a high frequency siren. The frequency used was about 20 kc, and its acoustic level, in the region where the subjects were placed, was between 160 and 165 db (relative to 10−16 watt/cm2). With sufficient exposure—from 10 seconds for flies and mosquitoes to 3 or 4 minutes for roaches and caterpillars—the sound proved lethal in all cases. More detailed work was performed on mice and the roach, Periplaneta americana. In both cases it was definitely established that the heating produced by sound absorption was sufficient to cause death. In addition to the heating there are other effects, notably tissue rupture, as evidenced by the almost complete distruction of the wings on flies and mosquitoes and the rapid deterioration and final disappearance of the external pinna of a mouse who had received a sublethal dose. During observations it has been impossible to completely avoid personal exposure to the sound field, and some of the effects observed under these conditions will be described. These included momentary dizziness, and heating of exposed parts of the hand.
20(1948); http://dx.doi.org/10.1121/1.1916917View Description Hide Description
Although a number of investigators have studied the summation of loudness of two or more tones heard simultaneously, there has been but little work on the summation of tones at or very near threshold. In the experiments here reported, summation effects at the masked thresholds for tones masked by random noise were investigated. The tones were studied in pairs, the separation of the two components being 25, 100, and 500 c.p.s., the frequencies being selected to maintain their geometric mean frequency at 800 c.p.s. The aural critical band width for binaural listening at this frequency is approximately 40 cycles; thus the tones selected were separated by slightly less than one critical band, slightly more than one critical band, and somewhat over ten critical band widths, respectively. The masked thresholds for the tones heard singly and in pairs were measured for five listeners for 60‐db and 20‐db masking. Addition of the second tone caused no significant improvement in recognition in any case except the one where the two tones were located within the same critical band. In this case a gain in recognition of approximately 1.7 db was obtained.
20(1948); http://dx.doi.org/10.1121/1.1916918View Description Hide Description
If the electrical impedance of an electromagnetic speaker is plotted as a vector when the acoustic loading consists of an air column of variable length terminated by a rigid plug, a circle diagram results. If the air column is terminated instead by an acoustic sample, a second circle diagram results. Various relationships between the two circles have been worked out which yield the acoustic impedance of the sample with considerable precision. The method inherently averages many individual readings and automatically checks the self‐consistency of the data. It is applicable to samples with little absorption since full advantage is taken of the possibility of coupling such a sample most effectively to the speaker by means of the air column. The method has been used to determine the acoustic impedance of commercial materials at low audiofrequencies, where the absorption is quite small.
20(1948); http://dx.doi.org/10.1121/1.1916919View Description Hide Description
The acoustic output impedance of a sound source can be determined from relative pressure measurements following a procedure which requires rather simple equipment and practically no time for processing data. The sound source must have a radiating area which is essentially a piston and which can be coupled to a closed tube of variable length. The tube is terminated by a rigid plug in which a pressure‐measuring microphone is mounted flush with the surface. Thus the microphonemeasures the pressure at the terminating plug as the length of the air column is varied. These pressure measurements are related to the acoustic impedance of the source by the expression , where p 0 is the zero length or blocked pressure,pl is the pressure at length, k is 2π divided by the wave‐length, l is the length of the air column, R is the specific acoustic resistance ratio for the source, and X is the specific acoustic reactance ratio for the source.
The principal value of the method lies in the fact that for any chosen length, the possible values of the pressure ratio represent a family of circles on acoustic impedance coordinates. Thus a chart can be very easily prepared from which the source impedance can be read and recorded as the data are taken. The circles can be labeled in decibels so that the only data reduction required is the subtraction of the pressure at length l in decibels from the blocked pressure in decibels. The author used the method in the “cut and try” development of a sound source for an acoustic filter experiment, where the dozens of impedance determinations would have represented a prohibitive amount of work if even moderately complicated data processing had been required.
20(1948); http://dx.doi.org/10.1121/1.1916920View Description Hide Description
Nichols and Wallace et al. have described aircraft acoustical treatments and apparatus for measuring the attenuation provided by them. This paper presents a theory and charts for calculating the attenuation of sound by six different multiple structures as a function of dimensions, specific flow resistance, density, and frequency. Five of the structures are typical of those used in aircraft, and two are types used in lightweight office partitions.
The calculations are checked against experimental data taken on a modified form of the apparatus of reference 2. The agreement between experiment and theory is satisfactory. Differences between attenuationsmeasured here and those described in reference 1 are explained.
20(1948); http://dx.doi.org/10.1121/1.1916921View Description Hide Description
A new procedure is presented for the measurement of the normal specific acoustic impedance of extremely porous screen as a function of frequency. The method (based on a special application of the transmission‐characteristic method of impedance measurement) is restricted to this case. Equations are derived which express the impedance of the screen in terms of the acoustical characteristics of a small enclosure, i.e., a rectangular chamber or a closed cylindrical pipe. Good accuracy can be obtained with relatively simple experimental equipment and with relatively little computation. Measured values of the impedance for samples of extremely porous cloth, determined by this procedure, are given.
20(1948); http://dx.doi.org/10.1121/1.1916922View Description Hide Description
A threshold type of noise reduction system has been developed for use in reducing the noise in sound reproducing systems. A non‐linear element allows the useful signal to pass but discriminates against the noise. This system lowers the ground noise level and thereby increases the signal‐to‐noise ratio without discrimination against the useful part of the signal. The operation of the system is instantaneous. There is no frequency discrimination for any signal level above the threshold.
20(1948); http://dx.doi.org/10.1121/1.1916923View Description Hide Description
It is well known that to obtain very small side lobe radiation, the amplitude distribution across a radiating surface should be such that the maximum amplitude is at the center and the minimum at the edges. The Gaussian distribution gives no side lobes but requires an infinitely large radiator. This paper gives the analysis of the clamped‐edge disk vibrating in its first normal mode. It shows that the dynamic curve approximates the Gaussian form and that the sound pressure distribution has very small side lobes, the amplitude of the first side lobe being 33 db below that of the axial lobe.
For a disk of a given diameter the width of the axial lobe may be decreased by raising the frequency of the first normal mode, which requires a corresponding increase in the disk thickness. At very high frequencies the disk thickness may become prohibitive. In such cases a thin disk can be forced to vibrate in the shape of the first normal mode by a proper radial distribution of the driving force.
20(1948); http://dx.doi.org/10.1121/1.1916924View Description Hide Description
Physical measurements of loudspeaker performance are becoming recognized increasingly as a reliable guide for the evaluation of loudspeaker excellence. The facilities of the Altec Lansing Corporation for the calibration of loudspeakers and microphones are described. The measurement techniques are illustrated with data describing a recent model loudspeaker showing the principal performance criteria, namely, frequency response, angular distribution, and distortion. Suggestions are given for a uniform presentation of performance data and specifications.
20(1948); http://dx.doi.org/10.1121/1.1916925View Description Hide Description
A new instrument was designed and built to record the subsonic atmospheric disturbances resulting from the explosive demolition of the German fortifications of Helgoland on April 18, 1947. Microbarometric disturbances flex the diaphragm of the microphone, altering one of two matched inductive circuits; the unbalance produces a signal voltage. This signal, the low frequency modulation of an audiofrequency carrier, is amplified and used to drive a modified Esterline‐Angus graphic recorder. To provide wide dynamic range as well as high sensitivity to small signals, the amplifier has linear response to low level signals and logarithmic response to high level signals. Completely linear response is available by switching out the automatic volume‐control circuit. Sensitivity may be reduced in steps to 1/5 or 1/125 to satisfy changing requirements, and at maximum sensitivity the recorder deflection is about 0.75 mm per microbar. As field equipment, the instrument system is necessarily battery operated. It is light and compact, and yet possesses high sensitivity and better stability characteristics than previous equipment of its type.
20(1948); http://dx.doi.org/10.1121/1.1916926View Description Hide Description
The first portion of this paper is a discussion of the design philosophy of a monitor speaker. Attention is paid to the apparent inconsistency between pressure measurements and A‐B tests with original material. Comment is made on the effects of change in directional pattern of the radiated sound. The performance of an experimental monitor loudspeaker is evaluated.
The second portion of the paper is a description of the experimental speaker. The configuration, significant dimensions, and motional impedance measurements are reviewed.
20(1948); http://dx.doi.org/10.1121/1.1916927View Description Hide Description
A discussion will be given of the purpose and aims of the Laboratory with relationship to the other underwater sound laboratories in the country, and a description of the various acoustic measuring systems and measuring techniques employed in the Laboratory. Included in the discussion of systems and techniques will be the high frequency system from 50–3500 kc, the low frequency system from 0–100 cycles (0–100 lb. p.s.i. and 34–100°F), the two intermediate frequency systems, 0.02 to 150 kc, and the pressure system, 10–150 kc at 0–300 lb. p.s.i.
Emphasis will be placed on the high frequency system and the high pressure system, and the methods of measurement under each system.
20(1948); http://dx.doi.org/10.1121/1.1916928View Description Hide Description
The directivity factor, the ratio of the average sound intensity from an electroacoustic transducer to the sound intensity in the reference direction, both measured at the same separation, is measured directly using the device described.
The equipment is composed of a mechanical mounting which permits rotation of the unit about two mutually perpendicular axes and electronic equipment which integrates the power. The electronic circuits used are of two different types; one type, designed by D. H. Ransom and J. H. Martin, uses a thermocouple, a direct‐current amplifier, a 60‐cycle vibrator, a 60‐cycle amplifier, and a modified watt‐hour meter; the other type, designed by R. C. Fisher and C. J. Burbank, uses the potentiometer in a power level recorder into which 60‐cycle alternating current has been introduced, a 60‐cycle amplifier, and a regular watt‐hour meter which has been modified to increase the dynamic range.
A measurement of the directivity factor and the input power to the transducer gives the over‐all efficiency of the unit in question.
The design considerations and use of the instruments are discussed.
20(1948); http://dx.doi.org/10.1121/1.1916929View Description Hide Description
Cavitation in the body of a liquid, throughout a region well removed from all boundary surfaces, has been produced by means of a converging sound wave. The cavitation phenomenon can thus be studied as a function of the state and properties of the liquid, without regard to the possible effects of minute irregularities or adsorbed gases on the surface of the sound generator. The method employed is to bring one or more sound beams of moderate intensity to a common focus within the liquid by means of a parabolic mirror. At the sixty‐kilocycle operating frequency of the magnetostriction transducers which serve as the projectors, attenuation in the liquid medium is small; nevertheless the use of a controlled resonant standing‐wave system permits the experiments to be conducted in a tank of reasonable laboratory size not equipped with any special energy absorbing devices. Moreover, the necessary driving power is reduced. The apparatus provides for measurement of the cavitation threshold of various liquids as a function of temperature, hydrostatic pressure, and pulse length. This paper will describe the equipment and present some preliminary results.
20(1948); http://dx.doi.org/10.1121/1.1916930View Description Hide Description
Measurement of the variation of sound intensity with range in El Capitan reservoir has been used to calculate the absorption of sound in fresh water. At 500 kc and 1000 kc the absorption coefficient is proportional to the square of the frequency, and when extrapolated to much higher frequencies is in approximate agreement with the values of Fox, and Fox and Rock for distilled water.Absorption at 120 kc was found to be too low for measurement within the limited range (1100 yards) of the reservoir, but an upper limit is established. Measurements in the ocean at 120 kc give a value which is in approximate agreement with those of the University of California Division of War Research; this value is more than five times the fresh water upper limit, and more than ten times the extrapolated higher frequency values. Measurements at 500 kc and 1000 kc indicate that oceanic absorption is not proportional to the square of the frequency; oceanic absorption at 1000 kc is only slightly higher than fresh water. It is suggested that the oceanic absorption mechanism has a relaxation time lying between 10−5 and 10−6 sec.
20(1948); http://dx.doi.org/10.1121/1.1916931View Description Hide Description
The attenuation of sound waves in distilled water has been measured in the frequency range from 50 to 450 kilocycles. The results are in good agreement with those of Fox and Rock in the range above 10 megacycles.
Measurements on filtered sea water from 50 to 180 kilocycles give values for the attenuation about 10 times as great as those obtained for distilled water. The measurements in sea water are in reasonable agreement with the measurements of other experimentors in the frequency range from 20 to 50 kilocycles.
The data to be presented were obtained by the reverberation method, using an improved spherical resonator.
20(1948); http://dx.doi.org/10.1121/1.1916932View Description Hide Description
The feasibility of conducting reciprocity calibrations up to 100 kc in air, using W.E. 640 AA condensermicrophones is pointed out, and the results of such calibrations will be presented. Curves of microphone sensitivity and the output of the unit when used as a sound source will be given for both a normal 640 AA unit and one whose diaphragm has been stretched to give increased high frequency response. In addition, curves for other transducers tested in this laboratory will be presented. All the calibrations were carried out in a small—6′×6′×6′—dead room.