Page 23 - Summer 2018

and the physical characteristics of orchestral music were published in a symposium of six papers in Transactions of the American Institute of Electrical Engineers (Affel et al., 1934; Bedell and Kerney, 1934; Fletcher, 1934; Scriven, 1934; Steinberg and Snow, 1934; Wente and Thuras, 1934; also see Jewett, 1933 for an overview).
To understand the requirements of any system that would re- produce the sounds of an orchestra in high fidelity, the Bell Telephone Laboratories team quantified the frequency trans- mission range needed to produce the sounds of each instru- ment in the symphonic orchestra without noticeable distor- tion to a skilled listener. The thin, wispy quality of a piccolo survived recording and playback so long as frequencies in the range of 500 to 10,000 Hz were preserved, but the pure sound of a violin required a wider bandwidth ranging from 40 to 13,000 Hz to transmit. This range was perhaps surprisingly broad; consider that the typical range of notes a violin can produce spans from 196 to 2,637 Hz (for more about violin acoustics, see Gough, 2016) . Of course, Fletcher’s team had the insight that clear reproduction of music wasn’t simply a matter of matching the highest and lowest notes that could be played from an instrument. Whenever a sound changed quickly, a theoretically infinitely wide bandwidth would be needed to convey that transition faithfully. Thus, the range of sounds that should be recorded and transmitted by an ideal system should match the human ear and not be tailored to the sounds that need to be reproduced (Fletcher, 1934). However, although those with normal hearing can, in theory, detect pure tones from 20 to 20,000 Hz, Fletcher reasoned that the extreme limits were not critical to preserve for his purposes: “In music these frequencies usually are at such low intensities that the elimination of frequencies below 40 cps and above 15,000 cps produces no detectable difference in the reproduc- tion of symphonic music” (Fletcher, 1934, p. 10).
With the same systematic approach, the researchers tackled the question of amplitude: what range of power must a sys- tem for reproducing orchestral music have to sound lifelike? Fletcher reported that “if this discussion were limited to the type of symphonic music that now is produced by large or- chestras,” a dynamic range of 70 dB would suffice (Fletcher, 1934, p.10). But he recognized that this limit was not imposed by any fundamental quality of music but more practically by “the capacities of the musical instruments now available and the man power that conveniently can be grouped together un- der one conductor.” Fletcher marveled at how the early pro- totypes of the Bell Telephone Laboratories engineers, which could both amplify and attenuate sound in recordings, en-
Figure 3. “Oscar,” a mannequin used for binaural recording experiments by Bell Telephone Laboratories personnel. Oscar was present for recording experiments during the 1931-1932 season at the Academy of Music. Reprinted from Rankovic and Allen (2009).
abled the Philadelphia Orchestra to surpass these technologi- cal barriers with aesthetically pleasing results.
As soon as a system was built that was capable of handling a much wider range, the musicians immediately took advan- tage of it to produce certain effects that they previously had tried to obtain with the orchestra alone, but without success because of the limited power of the instruments themselves (Fletcher, 1934, p. 10).
The Bell Telephone Laboratories researchers were also busy characterizing the physical requirements necessary for audi- tory perspective. They recognized two ways to produce this effect. One method involved head receivers, in which a dum- my head with two microphones positioned by the ears re- corded right- and left-hand audio tracks that could be played back to the listener over headphones. Fletcher and his col- leagues Snow and Hammer had worked extensively on this method of mannequin recording and had constructed “Os- car,” a department store mannequin with microphones in either cheek (for a review, see Paul, 2009) for both scientific study and demonstrations to the public (Figure 3). Indeed, Oscar was present in the audience at several of the Philadel- phia Orchestra rehearsals (Hammer and Snow, 1932)!
Summer 2018 | Acoustics Today | 21