Radio Broadcast (May 1928-Apr 1929)

BROADCAST ENGINEERING BY CARL DREHER Pick-Up Characteristics of Microphones CONSIDERABLE work is being done in determining the behavior of microphones under various conditions of studio pick-up. While the results will probably not be reported for months or years, the literature already contains some material of practical value for those who are interested in securing the best possible quality of reproduction as well as for laboratory technicians who use microphones in sound-measurement determinations. B. F. Miessner's article in the September, 1926, Radio Broadcast, on "The Importance of Acoustics in Broadcasting" is worth rereading in this connection. Miessner was concerned in this paper with possible distortion in radio reproduction caused by the directional characteristics of microphones and loud speakers. He concluded that these devices usually vary in directional characteristics with frequency. For horn speakers and flat diaphragms enclosed on one side he secured a polar diagram, reproduced herewith as Fig. 1, which shows a regular falling off in intensity from front to rear at low frequencies, the presence of a minimum at 90-120 degrees at higher frequencies, and a marked beam effect at still higher frequencies. This beam effect was also very noticeable with cone and baffleboard-type loud speakers, as well as the dynamic units. Miessner argued that such an effect as that of Fig. 1, secured by measurements on a horn or diaphragm of about 12" diameter, would also be noted in pick-up work with the same device, the action being a reversible one. "It is plainly evident," he wrote, "that if a musical instrument, say a 'cello with low-pitched fundamental and high-pitched overtones, be placed at an angle of 45 degrees to the face, as it well might in a studio, the fundamental would be received about 75 per cent, as loud as if it were in front of the microphone, while overtones of the order of 5000 cycles would be reduced to less than 10 per cent." He went on to raise the point that a square-law effect might be involved when the directional distortion of the microphone is repeated by the loud speaker. While this is true, quantitatively Miessner's illustration of the 'cello is somewhat misleading under practical conditions, as he himself recognizes, for toward the end of the article he modifies his conclusions as applied to the then standard broadcasting microphone of the Western Electric 373-w double-button carbon type, now superseded by the 387-w. Although this microphone has a closed back, it responds to sounds from the rear because of diffraction around the housing. The facility with which the sound wave bends around the obstruction depends on the wavelength compared to the size of the obstacle. If the microphone housing is small compared to the wavelength, diffraction takes place with little loss in intensity. For higher frequencies, on the other hand, the diaphragm may be in a region of pronounced acoustic shadow, resulting in discrimination against high notes. With an actual microphone diaphragm and housing the ratio of dimensions to wavelength is not as unfavorable as Miessner's curves of Fig. 1 would indicate, and he gives another set of polar diagrams, (Fig. 7 in the original article) here reproduced as Fig. 2, which approximate actual broadcast pick-up conditions. In the latter, it will be noted, the discrimination at 45 degrees against a 5000-cycle tone, compared to a 100-cycle fundamental, is not of the order of 7.5, but only about 2.3. A simple expedient used by broadcast engineers in order to reduce loss of the high frequencies in picking up music over a wide front, as in the case of an orchestra of good size, is to employ two microphones facing outwards at right angles (Fig. 3) mounted on a single stand a few inches apart. This doubles the angle in which pick-up occurs without serious directional distortion. If this angle is 90 degrees for each transmitter, the two will cover a total of 180 degrees, or all of the space in front of the microphone stand. The outputs of the two microphones are mixed in the usual way (Fig. 4) where the repeating coils have 200-ohm windings to match the impedance of the microphones, and the potentiometers are about 400 ohms each, the combination working into the 200-ohm input of the amplifier. An additional advantage of such a combination lies in the fact that pick-up is not confined to one point in the room and there is less chance of running into any serious acoustic anomalies arising from interference of reflected waves or other effects of the room characteristics. However, a right-angle microphone combination of this type presents no advantage in picking up announcements or other close-talking material. THE BEAM EFFECT The beam effect of projection of high frequencies is well recognized now in human articulation, the output of many musical instruments, and in loud speaker design. Pick-up of ordinary speech with present-day equipment is generally defective when the speaker is not talking directly into the microphone because the high frequencies, which are so important in the interpretation of speech, issue in a beam in the direction in which the Fig. 1 — (left) Polar curves showing directional characteristics of horns and flat diaphragms enclosed on one side at frequencies of (1) 100, (2) 1000, and (3) 5000 cycles Fig. 2 — (right) Approximate directional characteristics for broadcast microphones at standard test frequencies o f 100, 1000, and 5000 cycles, respectively 220 140 150 210 160 200 170 190 190 170 200 160 210 150 • march, 1929 . . . page 311 •