Volume 24, Issue 4, July 1952
Index of content:
- PROGRAM OF THE FORTY‐THIRD MEETING OF THE ACOUSTICAL SOCIETY OF AMERICA
- Session A: Noise and Noise Control
- Contributed Papers
24(1952); http://dx.doi.org/10.1121/1.1917467View Description Hide Description
The conventional methods of design of axial flow compressors usually consider the losses introduced by the generation of noise to be of negligible magnitudes and therefore ignore them entirely in a study of compressors. As a result, there is scant information on the relation between the noise level and the performance properties of compressors. In the present paper an attempt is made at working out the basis of a method of evaluation of the noise characteristics from compressors. The method makes considerable use of the experimental findings of Beranek and Rudmose on noise due to airplane propellers. The analysis presented here considers the noise to be generated by a number of point sources of equal magnitudes but of random phase. The resultant intensity level of these point sources (presumed to be located at the tip of the compressor blades) is calculated on a statistical basis. The expected noise level is given as a function of blade tip speed, clearance between blades and stator case, number of blades and the power expended in driving the compressor.
24(1952); http://dx.doi.org/10.1121/1.1917468View Description Hide Description
Noise resulting from the flow of air over the fuselage of an airplane has been investigated and found to depend principally on approximately the fifth power of the indicated air speed. The resulting emperical relation of noise level to airspeed is found to be quite useful in estimating the noise in the frequency range above 600 cps, for airplanes varying widely in size, shape, and method of propulsion. From this relation a method has also been evolved for making rough estimates of the weight of acoustical insulation necessary to reduce the noise in airplanes to a predetermined level.
24(1952); http://dx.doi.org/10.1121/1.1917469View Description Hide Description
Much of the lack of correlation between noise level and hearing damage is because of unsatisfactory theories which express the effect of frequency. It is possible to get a simple damage risk criterion, if one makes the following assumptions: (1) that hearing damage for those who work in noisy environment is a progressive auditory fatigue phenomenon, (2) that the fatigue is the same function of the stimulus in the inner ear for all frequencies, and (3) the sensation level is a similar function of the stimulus in the inner ear for all frequencies. The damage is, on this hypothesis, simply a function of the sensation level and varies with frequency in the same manner. Fletcher's calculation of the energy levels in the fluid near the basilar membrane as a function of sound pressure and frequency supports this theory. This analysis leads to the setting up of equal loudness contours for broad‐band noise which correspond to hearing damage contours. Such curves explain for the first time the 4000 cps “notch” in the audiograms of industrial workers. Research data taken from a survey of industrial noise serve to establish two octave band curves which tentatively appear to bracket the region for damage risk for daily exposure over many years to a steady state noise environment. When the noise exceeds 100 sones in any octave band, long exposure definitely appears to be damaging. When the noise does not exceed 50 sones in any octave band, there appears to be very little danger of hearing damage even after many years of working day exposure. These conclusions are supported by the few data available in the literature. Evaluation of clinical data, which have not yet been disclosed by industrial audiologists and otologists, should make it possible to confirm and more sharply define the damage risk levels.
24(1952); http://dx.doi.org/10.1121/1.1917470View Description Hide Description
Presented in this paper is a set of criteria that have been developed and applied on numerous noise control problems. A wide range of confirmatory evidence and experiences has led us to view these criteria with some confidence, although we recognize that they are subject to modification on the basis of additional data which are continuing to accumulate. An essential feature of the criteria is that they incorporate a specification of frequency dependence (usually expressed by octave bands) as well as over‐all sound pressure level. Tentative criterion curves are presented for: (a) risk of permanent damage to the hearing mechanism under habitual exposure to (i) continuous spectrum noises and (ii) single frequency components; (b) speech communication conditions in terms of percentage intelligibility, voice level, distance between talkers, and type of vocabulary; (c) risk of annoyance to residential areas as dependent on type and previous noise conditioning of the community, daytime vs nighttime exposure, and the time pattern of the exposure. Emphasis is laid on the importance of interpreting these criteria in the light of the inherent variability among individuals in their responses to noise stimuli.
24(1952); http://dx.doi.org/10.1121/1.1917471View Description Hide Description
In many jet engine testing facilities, water sprays are installed in the exhaust piping to cool the exhaust gases to temperatures which will not damage the walls of the pipe. The water sprays appreciably alter the character of the sound propagation through the piping. The present paper discusses the dependence of the attenuation of sound on such characteristic parameters of the water‐spray muffler as its geometry, the amount of air circulated, the quantity of water sprayed, and the average sizes of the waterdroplets in the air stream. The attenuation of sound in the muffler is found to be due mainly to viscous losses in the boundary layer between waterdroplets and air. The theory discussed demonstrates that in order to achieve maximum sound attenuation through such a muffler there are optimum conditions as to quantity of water used and the method of its injection into the air stream. The relative velocity of the air stream to that of the injected water must lie within a critical range in order to fix the average size of the waterdroplets.
24(1952); http://dx.doi.org/10.1121/1.1917472View Description Hide Description
An attempt has been made to obtain reliable information on the intrinsic performance of sound attenuatingtreatments used in the air flow passages (stacks, ducts, tunnels) of test cells and other aircraft engine test facilities. The apparent performance in a particular installation may be influenced by flanking transmission through the surrounding structure, by standing waves within the cell, by reflections at the ends of the treated section, by the radiation pattern at the exit, and by the spectral characteristics of the noise source for the measurements. These effects must be assessed separately if one is to obtain valid data on the attenuation value of the treatment itself. One can then correlate measurements from many installations and from laboratory experiments to yield generalized performance characteristics. Some results so obtained are given for (a) uniformly spaced absorptive baffles within and parallel to the air stream over a range of baffle thicknesses and separations, (b) absorptively lined sections of ducts, and (c) right‐angle bends with absorptive lining facing the incident sound. Illustrative of the generalization, the data on bends are fitted to a single curve plotted against the ratio of duct width to wavelength with good agreement where the absorptive lining satisfies certain scaling restrictions.
24(1952); http://dx.doi.org/10.1121/1.1917473View Description Hide Description
The noise produced by a large experimental supersonic windtunnel in which a shock wave was employed to produce high wind velocities and in which a jet engine was operating was sufficient to disturb residents of Cleveland within five miles during after‐midnight operations early in 1950. The disturbing noise extended from 5 cps to 8000 cps. Components between 4 and 20 cps were reduced by a series of resonators. Components between 20 and 700 cps were reduced by a set of six parallel lined ducts properly chosen in size and lined according to charts of P. M. Morse. Above 700 cps, the reduction was provided by a series of parallel absorbent baffles and two lined bends. The resultant acoustic muffler renders the noise unobjectionable even in nearby office buildings, without affecting significantly the aerodynamic performance of the windtunnel.
- Session B. Psychoacoustics and Hearing
24(1952); http://dx.doi.org/10.1121/1.1917475View Description Hide Description
What Stevens has called the electrophonic effect, i.e., the sensation of hearing due to an electric current passed through the head, was first observed by Volta in 1800, and for many years it was thought to be a single phenomenon having a single cause. Jones, Stevens, and Lurie showed that at least three different kinds of sensation, involving three separate mechanisms, may result from electric currents applied to the head, but it now appears that electrophonic hearing is even more complicated than they supposed. Depending upon the kind of electrode system employed, a sinusoidal electric current gives rise to at least five phenomena. (1) Electrode immersed in salt solution in the ear (the most common method). The subject hears a complex tone, containing mostly second harmonic, suggesting the action of a square‐law transducer. (2) Electrode in contact with the epidermis of the meatus. The subject hears the first harmonic, and at low frequencies he may also hear a noise. (3) Electrode on the mucous tissue inside the middle ear (with the eardrum removed). The subject hears the first harmonic and/or a noise. (4) Large‐area electrode on the skin, any place on the head. Provided the skin is dry the subject hears the second harmonic; if it is wet, he hears nothing. (5) Moving electrode on the skin or on the roof of the mouth (fricative effect). This gives the strongest sensation of hearing. The subject hears either the first harmonic or the first and second harmonics, or the second harmonic alone, depending upon the applied voltage and upon the placement and the properties of the electrode. The experimental results seem to indicate that the hearing of a tone under any of these five conditions is probably due to vibrations set up outside the cochlea, although there appears to be at least four different transducing mechanisms. Contrary to earlier hypotheses the tympanic membrane is apparently not involved in the conversion of the electrical energy into mechanical vibration.
24(1952); http://dx.doi.org/10.1121/1.1917476View Description Hide Description
Electrophysiological study of the ototoxic effects of streptomycin in the cat has shown that prolonged treatment with large doses of the antibiotic may cause varying degrees of depression of the microphonic response of the cochlea, with or without comparable reduction in the amplitude of the response of the auditory nerve. Hydroxystreptomycin exhibits this specific toxicity in a much higher degree, while certain preparations of neomycin cause rapid and complete loss of the functions of both the microphonic mechanism and of the nerve. Commercial dihydrostreptomycin, on the other hand, usually causes an enhancement of the microphonic response and, at the same time, depression and splintering of the action potential patterns of the nerve. The nature of the ototoxic action and its significance for the study of auditory physiology are discussed.
24(1952); http://dx.doi.org/10.1121/1.1917477View Description Hide Description
On the basis of experience with previous models, new ear wardens have been developed in an effort to improve the intelllgibility of speech heard through the wardens and the protection of the hearing of the wearer, and to provide him with maximum comfort. One of these wardens attenuates the whole frequency range important for hearing, one serves as a low pass filter. The choice between them depends upon the characteristics of the noise encountered and the activity of the wearer. Experiments have been carried out to assess the protection afforded under a variety of conditions in both industrial and military situations.
24(1952); http://dx.doi.org/10.1121/1.1917478View Description Hide Description
The experience gained in fitting a large number of impaired persons with hearing aids is presented, showing illustrative ranges of impairment between 30 and 100 decibels fitted with three progressively powered hearing aids. General features of each hearing aid are discussed, showing the circuits and arrangements for peak clipping and automatic gain control. The microphone, air and bone receivers are described, considering features developed for providing optimum reliability in service, such as moistureproofing and design arrangements providing miniaturization concomitant with ruggedness.
24(1952); http://dx.doi.org/10.1121/1.1917479View Description Hide Description
Audiometric tests (pure tones and speech) were conducted on deafened ears in quiet and at several levels of calibrated white noise. The cases were classed into six groups based upon results of their threshold audiograms. “Recession” describes the gradual approach toward normal levels with increasing noise of masking curves obtained from such ears. In cases displaying recruitment recession was relatively rapid, although not complete in some cases. A pure tone‐to‐speech perception ratio was established between the three audiometer frequencies within the speech range and the level of the 50 percent spondee score. Corrections were applied for (1) the type of loss and (2) the subject's unfamiliarity with the quietness of audiometric test rooms. This ratio is expected to differentiate speech losses as to their auditory or extra‐auditory origin. From the results of speech tests, the “Social Adequacy Index for Hearing” in noise was evaluated in the same manner customarily used under quiet conditions. Adjustment of the test results was necessitated because of the narrowing of the “Useful Auditory Area” induced by noise. The SAI frequently increased but in some cases decreased with noise, indicating the relative hearing efficiency in noise of deafened subjects.
24(1952); http://dx.doi.org/10.1121/1.1917480View Description Hide Description
This paper describes a method of calculating the articulation scores which a talker‐listener pair will obtain in terms of the response of the hearing aid, the audiogram (both bone and air) of the listener, and talking level of the talker, and the proficiency factor of the talker‐listener pair. This method was used to calculate the articulation scores for all the talker‐listener pairs and the six kinds of hearing aids used in the test at Harvard made by Davis and his associates. The calculated and observed results agree within the experimental error. The philosophy underlying the method of calculation is used to predict the ideal type of response to be used by a deafened listener. It leads to the following simple rule. The ideal response R in db gain at any frequency is given by , where β A is the hearing loss by air conduction, β B is hearing loss by bone conduction, and α I is the over‐all gain (same for all frequencies) to bring the speech level at the listener's ear to its optimum value. If the values of at the three frequencies 500, 1000, and 2000 chs are examined and the average of the two highest values called β2, then the maximum value of α I is given by . The optimum gain α I is, in general, less than α I (max), and so a means of lowering the gain in the set must be provided. When α I (max) calculates to be negative, then the optimum gain α I is equal to α I (max). These gains are all with reference to a speaker 100 cm from the listener, and speaking so that his average level at the listener's ear is 66 db. If the speaker moves farther away or talks softer, then it is counted a negative gain or a loss. If he moves closer or talks louder it is counted as a gain.
24(1952); http://dx.doi.org/10.1121/1.1917481View Description Hide Description
A question is raised regarding the operation of the two time constants in hearing. The short‐time constant of about 15 milliseconds, due presumably to the cochlear filter system, is a “signal amplitude integrator.” The long‐time constant of about 0.5 second, due presumably to the central nervous system, is a “signal power integrator.” Under some conditions not yet understood, the short‐time constant can be directly observed by signal duration experiments without interference from the long‐time constant. Some experiments were performed on the detection of pure tones in white noise in which both “signal detections” and “false alarms” were recorded. It was found that the listener does not treat errors of the “signal miss” type in the same manner as he treats errors of the “false alarm” type. For signal detectabilitles ranging from 30 to 90 percent a typical observer will have an almost imperceptable false alarm probability, usually much less than 1 percent. This result will be discussed in the light of the theoretical behavior of the two integration mechanisms mentioned above.
24(1952); http://dx.doi.org/10.1121/1.1917482View Description Hide Description
The unexplained difference in sound pressure in the ear canal which appears to exist when equally loud low frequency tones are presented alternately from an earphone and from a loudspeaker has bedeviled acousticians for many years and, unfortunately, still continues to do so. There are presented here the results of some of the measurements carried out at the Bell Telephone Laboratories which show the magnitude of the effect and various attempts at explaining it. While no satisfactory explanation has been found, it is hoped that publication of these results will stimulate interest in the problem.
24(1952); http://dx.doi.org/10.1121/1.1917483View Description Hide Description
During a current investigation of ears with chronic high frequency tonal gaps, it was found that some sort of uniaural diplacusis is usually associated with this type of hearing abnormality. Generally this uniaural diplacusis takes the form of a “rough” or “noisy” sensation within certain frequency ranges. In one subject, however, the distortion is definitely tonal in character; in particular, frequencies from 2700 to 3300 and from 3900 to 4500 cps are heard as two or more separate tones. Extensive tests show that when an objective frequency F (within the above ranges and below 60 db SPL) is presented, one or more of the following tones is heard in addition to F: 2F‐3600, 2(3600)‐F, 3F‐2(3600), 3(3600)‐2F. (These tones represent the difference tones heard in a normal ear when F and an objective tone of 3600 cps are presented simultaneously.) Also, rough beats are heard at threshold for frequencies from 3500 to 3700, with a null point at 3600 cps. Finally, when the acoustic stimulation consists of alternate one‐second tones and silence, the silent period is filled with a 3600‐cps “after‐clang.” These results suggest some basic instability of this ear at 3600 cps. There is, however, no persistent tinnitus. The similarity of the observed effects to the “personal tone” reported by Flottorp and to the “tinnitus beats” of Wegel is discussed.
24(1952); http://dx.doi.org/10.1121/1.1917484View Description Hide Description
Some measurements have been made on attenuation in bone in order to provide basic information needed in certain medical applications of ultrasonics. Valid data are difficult to obtain because of inherent variability in bone thickness and structure, and because adequately fresh bone samples are not easy to obtain. A special “bone gun” was used to obtain disks from cadaver skulls, and measurements were made on some 35 of these disks within a few hours after death. Attenuation coefficients (in db/cm) were determined by measuring the insertion loss of samples of different thickness. The attenuation increases by more than 60 db/cm over the range from 0.26 to 3.5 megacycles, but the function is not linear. A possible mechanism for the nonlinear behavior is suggested.
- Session D. Psychoacoustics and Speech Communication
24(1952); http://dx.doi.org/10.1121/1.1917486View Description Hide Description
Whereas the ear's sensitivity for detecting a difference in frequency between two tones is remarkably acute, the ability of listeners to identify (and name) tones presented in isolation is relatively poor. When the frequency of a single tone (in the frequency range 100–8000 cps) is varied in equal logarithmic steps (and when the sound level is arbitrarily varied to reduce loudness cues), the amount of information transferred is about 2.3 bits. (The equivalent proficiency of response for an informational transfer of 2.3 bits is perfect identification among only 5 tones.) The amount of information that can be transferred is, within rather wide limits, independent of the number of tones and the range of frequencies employed.
- Session E. General Acoustics and Transducers
The Acoustics Design of the Large Council Chambers, Permanent Headquarters of the United Nations, New York24(1952); http://dx.doi.org/10.1121/1.1917493View Description Hide Description
The design of large parliamentary chambers for conferences conducted in many languages presents a number of new acoustics design problems. Not only must the acoustic environment be “comfortable,” but it must be possible to hear clearly a person speaking in any part of the delegates area, either directly or as simultaneously translated into any of four or five languages. The design and isolation of the many booths for translators, radio, and press surrounding these rooms was the subject of much study. The general isolation of noise from surrounding service areas and outdoor sources, such as traffic on the East River and in the air, were also considered. Special attention was given to the coordination of the sound reinforcing system with the acoustic design of the chambers. Preliminary articulation tests have been conducted in these rooms and the results are presented.
24(1952); http://dx.doi.org/10.1121/1.1917494View Description Hide Description
The problem of the reflection of acoustical waves from an infinite wall, having prescribed irregularities or a prescribed distribution of acoustical material upon it, is formulated in terms of an integral equation. By taking the Fourier transform of the equation it can be reduced to a form easily handled by numerical techniques. Expressions for scattering and absorption cross sections can be obtained without actually solving the equation. Much information of a qualitative nature can be obtained from inspection of the equation. An “index of diffusion” is introduced which is a measure of the effectiveness of a given surface in breaking up the reflected wave. The application of the above methods to simple scattering surfaces (e.g., strips and pillars) will bez discussed.