Binaural Receivers.—The principle of binaural listening ex pounded by Rayleigh has been extended to the use of artificial "ears" which can be made very sensitive to a given type of sound and can be arranged at any convenient distance apart or with any orientation with respect to the source of sound. The prin ciple has been applied to aeroplane and to submarine location (see Mechanical Properties of Fluids—" Submarine Listening," Drysdale). For example if two similar receiving trumpets are connected by equal lengths of tube to each ear, the trumpets being mounted a fixed distance apart on a rotatable platform, it is possible to locate a source of sound such as an aeroplane with considerable accuracy. The trumpets amplify the sound re ceived by the ears and the directional accuracy is increased by increasing the distance apart of the trumpets. As the platform is swung round, the path-difference of the sound reaching the two ears diminishes on one side and then increases on the other, giving the impression of the sound crossing from one ear to the other. At the point of "cross-over ", i.e., of binaural balance, the line joining the trumpets is at right angles to the direction of the source. The same principle can be applied to a pair of under water receivers. As an alternative to rotation of the line of the receivers it is often more convenient to vary the path-length to the two ears, the difference of path, relative to the distance apart of the receivers, being a measure of the "rotation " required to bring them symmetrical with respect to the source. This is known as the binaural compensator principle. The principle has been extended to multiple receivers, that is, to a considerable number of similar receivers equally spaced on a long base line. Instead of rotating the whole base line, a suitable rotary compensator is provided, the direction being indicated at the point of binaural balance. The directional properties of multiple receivers are the same as for corresponding multiple sources (see In the case of electrical receivers the binaural method may be replaced by the Sum and Difference Method—the effects at the ear being due to the receivers assisting or opposing each other. The sound received by the ear is a maximum or a minimum respectively, when the normal to the line of the receivers is in the direction of the source of sound.
Receivers of Large Area.—Just as a directional source of sound is obtained by using a vibrating surface of large dimensions com pared with a wave-length, so a receiver becomes increasingly directional as its dimensions increase. The piezo-electric quartz directional transmitter used by Boyle and Langevin serves equally well as a directional receiver. Similarly, large trumpets, mirrors and other focusing devices exhibit directional properties with increasing size or with increase in the pitch of the sound (i.e., with diminishing wave-length) Reference has already been made to the reflection which takes place at the open end of a cylindrical pipe or a trumpet acting as a sound receiver. The extent of this reflection diminishes, and the trumpet becomes a more efficient receiver, as the size of the opening increases, pro vided the length of the trumpet is always large. A parabolic mirror converges the sound energy towards a focus, but it must not be assumed that this is in any way an "optical focus," al though it approximates more and more to it as the wave-length of the sound diminishes. A watch tick, the sharp crack of a pistol or an electric spark may be focused very effectively by a mirror of moderate dimensions. The increase of loudness due to curv ature depends on the area of the reflecting surface, from which disturbances of uniform phase arrive, as compared with the area of the first Fresnel's zone of a plane reflector in the same posi tion. If the "focal length" of the reflector is considerable and X is not small, the first Fresnel zone is fairly large, therefore for a reflector of moderate dimensions there is little to be gained by making it concave.
Diffraction Disc Method.—On p. 26 reference was made to the fact that the sound intensity at a point on the axis of a circular obstacle placed normal to a beam of sound is as great as if the obstacle were not present, i.e., the sound "shadow " of the disc has, to quote the optical analogy, a bright spot at the centre. A microphone, tuned to the frequency of the incident sound, and placed at a suitable point on the axis of such a disc forms a very sensitive directional receiver, the response being a maximum when the sound is normal to the disc. Discs 20 feet or more in diameter have thus been used in combination with resonant hot wire microphones as an extremely accurate and sensitive means of locating distant sources of sound in air.
The location of sounds of an explosive nature by the use of a number of receivers on measured base lines is described in the section on Sound-Ranging (see p. 35).
reliable information in regard to the frequencies of the tones present in the sound-wave gives only approximate information regarding the relative intensity of these tones.
Numerous instruments have been devised which are described as intensity or sound-energy-meters, many of which give indica tions in some way related to sound-intensity. In no case, how ever, is the device universally applicable to all frequencies and intensities. Many attempts have been made to record wave forms of sound by means of diaphragms and stretched mem branes. The movements of a vibrating diaphragm excited by sound-waves have been recorded in various ways, mechanically, optically and electrically. Mechanical recording is exemplified in the gramophone, where a needle attached via a lever mech anism to the diaphragm traces the wave-form on a revolving disc or cylinder. (See GRAMOPHONE.) Optical methods have been very commonly used, e.g., in Kenelly's analysis of the motion of telephone diaphragms, in Webster's phonometer, in Millar's phonodeik and in the Hilger audiometer. In all these examples a small mirror reflects the vibrations of the diaphragm on to a moving photographic film. Electrical methods involve the use of a microphonic or electro-magnetic device which responds to the vibrations of the diaphragm, the corresponding electrical oscilla tions being recorded by means of some form of oscillograph, e.g., the Duddell strip, the Einthoven string or the cathode ray oscillograph (see INSTRUMENTS, ELECTRICAL; see also Irwin, Oscillographs). The majority of such diaphragm receivers are subject to resonances at certain frequencies with the result that the record is a distortion of the actual wave-form, the dis tortion being greatest in the neighbourhood of the resonance fre quencies of the instrument. The defect may be exaggerated or partially compensated by the use of a cone or trumpet to collect the sound and increase the amplitude of the diaphragm. Great care is necessary in the choice of a receiver, and recorder, for a particular type of sound, otherwise a number of factors may con tribute to the distortion. Thus in electrical recording, distortion may be caused by resonance in the diaphragm, the microphone mounting, the trumpet or cone collector, the electrical circuit, and in the oscillograph. In addition to this, there may be dis tortion introduced solely on account of the dimensions of the receiver relative to the wave-lengths of the various components of the sound. As we have seen, a receiver which is large compared with a wave-length exhibits directional properties depending on the ratio X/D (where D is the diameter of the receiver). One of the best non-resonant receivers developed during recent years is the stretched diaphragm condenser microphone of Wente (Phys. Rev. 1o. 39. 1917). This was specially designed for sound-measure ments and is used extensively in analysis and recording of wave forms of sounds lying within the frequency ranges of speech and music. The microphone, shown in diagram fig. 16a, consists essentially of a tightly stretched thin steel diaphragm (•ooi in. thick) separated from a parallel back-plate by an air gap of o•ooi in. approximately. The diaphragm shown with its back-plate form an electrical condenser, the capacity of which is varied when vibration takes place. When 200 volts are applied through a high resistance to the condenser, the vibration results in a fluctuating electromotive force. The variation of sensitiveness of such a microphone with frequency is indicated in fig. 16 b. It will be seen that the output in milli volts volt) per dyne/ cm.' is fairly constant over a frequency range 500 to 5,000 cycles/ sec. The microphone is very insensitive when compared with other, more familiar, types (granular and electro-magnetic) but its freedom from resonance over a wide range of frequency renders it most valuable for purposes of sound analysis. The lack of sensitiveness can easily be remedied by the use of a "distor tionless " valve-amplifier (resistance-capacity coupled valves). Crandall (Bell. Syst. Tech. Jour., 4. 1925, p. 587) employed Wente's condenser microphone in conjunction with such a valve amplifier to record speech sounds. A well designed amplifier will give practically constant amplification over the range of speech frequencies and will give an output proportional to input. The amplified e.m.f. is recorded by means of an oscillograph which must itself be non-resonant over the range of frequencies to be recorded. With certain limitations a high frequency Duddell strip well damped may serve the purpose, or alternatively an Einthoven oscillograph with a critically damped fibre of silvered quartz may be used. Perhaps the most perfect form of sound recording system is one proposed by Sir J. J. Thomson, namely a piezo-electric crystal receiver used in conjunction with a cathode ray oscillograph. Such a combination was used by Keys (Phil. Mag., Vol. 42, 1921) in recording the pressure-time curve of an explosion-wave under water. The piezo-electric receiver has a very high natural frequency and is consequently non-resonant over a very wide frequency-interval which includes the audible range. The cathode ray oscillograph (see A. B. Wood, Proc. Inst. Elect. Eng., Nov., 1925) is a perfect non resonant recorder of electrical oscillations, having the same sen sitivity at all frequencies from zero to the highest "radio" fre quency. The combination may therefore be regarded as distor tionless. The sensitivity for sounds of moderate intensity is, however, very small and amplification is necessary. The faith fulness of reproduction is ultimately dependent on limitations of the amplifier.