Radio broadcast .. (1922-30)

108 the studio microphone, assuming all impedances to be matched so that the line and studio micro- phone feed the first control room stage of ampli- fication with equal efficiency? (Solution on page 109) Design and Operation of Broad- casting Stations 16. Studio Design A LARGE part of the material in this dis- cussion is taken from Bureau of Stand- ards Circular No. 300, Architectural Acoustics, obtainable from the Superintendent of Documents, Government Printing Office, Washington, D. C., at five cents a copy. This pamphlet in turn leans heavily, as do all works on the subject, on the investigations of the late Wallace Clement Sabine, now being carried on by P. E. Sabine. These studies are concerned with the acoustics of auditoriums, but the results are adaptable to studios of broadcasting stations on the basis that the best reverberation time for electrical reproduction is about half that of an auditorium of the same volume where the per- formance is intended only for the audience physically present. The principal acoustic characteristic of a room is its reverberation time, which, as originally defined and measured by Sabine, was the time required for a sound a million times audibility to die down to just audibility, both expressed in power units. In other words, you take a sound of a certain intensity, and let it die down 60 TU (in telephone terminology) in an enclosure, and the time you have to wait is the reverberation time of that enclosure. Actually, reverberation times are generally calculated, and their prin- cipal use is comparative and empirical. For example, you have a certain studio whose char- acteristics have been found to be satisfactory. Then, if you have occasion to design another studio, it is wise to calculate the reverberation time of the first room and duplicate it in the second, or to make such modifications as are indicated by past performance. Sound may be reflected by the walls, floor, and ceiling of a room, or absorbed, depending on the materials of which the surfaces are composed, the frequency, etc. Generally, there is partial reflection and partial absorption. Open space is taken as the perfect absorbing material; thus, in a room, open windows are perfect absorbing areas. If a square foot of open window is taken as the unit of perfect absorption, various coeffi- cients may be applied to square feet of wall materials to indicate their power of absorption. A few such absorption coefficients are given below: In many cases these coefficients vary with fre- quency; theory indicates, in fact, that a flat characteristic for a homogeneous absorbing material is impossible. The coefficients given are supposed to be correct for 512 cycles per second, which is a good mean musical frequency, as 800 cycles per second is a good mean speech frequency for commercial telephone calculations. As would be expected, hard materials like marble, plaster, etc., absorb sound very little, whereas RADIO BROADCAST curtains, felt, human beings, etc. have high absorbing power. This is borne out by practical observation; everyone has noticed the persist- ence of sound in empty rooms and auditoriums, especially where the walls are hard and unbroken, while when curtains, furniture, and persons are present, reverberation is reduced. As the volume of the enclosure increases, the reverberation time also increases. Given the volume of a room, and the areas and absorption coefficients of the materials used in its construction, the reverbera- tion time may be approximately calculated by the empirical formula: o.o; V A (0 where / is the calculated reverberation time in seconds, V the volume of the room in cubic feet, and A the total absorption. We find A by taking the area of the various materials in the room and multiplying each by its coefficient of absorption, then adding the quantities so obtained. Obviously this assumes that the ab- sorbing power of a surface depends only on its area and acoustic characteristics, and is inde- pendent of its position in the room. This is sub- stantially true. The application of the formula is most readily learned by the actual solution of a problem: Given a room 20 by 15 by 10 feet, with plaster walls and ceiling, and a varnished hardwood floor, calculate the reverberation time, and apply acoustic treatment to reduce this period to an allowable value for broadcast purposes. V equals 20 times 15 by 10, or 3000 cubic feet. Let W represent the wall area, C the ceiling area, and F the floor area. Then we find in the above case that #^=700 square feet, C=3oo square feet, and F=3Oo square feet. A may now be calculated for the bare room: Wall plus ceiling area (lopo sq. ft.) multi- plied by the coefficient for plaster (0.025) 25 Floor (300 sq. ft.) multiplied by the coeffi- cient for hardwood (0.03) .... 9 Total absorption (A) ^4 Substituting in Formula (i) above, we have: o.o; (3000) 34 = 4.4 seconds This is much too high for a small room. Sabine found that good musical taste required a rever- beration time of slightly over one second, for piano music, in a room of moderate size. Musi- cians check each other with fair accuracy in such determinations. Even a large concert hall will have an optimum reverberation time of between two and three seconds only. (The allowable reverberation time increases approximately as the first power of the linear dimensions of the room, for the same absorbing materials, since V in the numerator of formula (i) is a cube function of the linear dimensions, while A, in the denominator, varies as the square of the linear dimensions.) Hence it is necessary to apply additional absorption to the room we are considering, in order to reduce the reverberation time to about 0.5 second. This is on the theory that the optimum period would be about i.o second, for a room of this size, considering results in the room only, and that this quantity should be halved for purposes of electrical reproduction at a distance. Suppose, now, that the ceiling of our 20 by 15 by 10 room be covered with some acoustic board material, or hair felt under muslin, with a co- efficient of, say, 0.5, and that the floor is car- peted, the walls remaining plaster. The total absorption, A, then becomes: JUNE, 1927 Ceiling area (300 sq. ft.) multiplied by 0-5 150. Floor area (300 sq. ft.) multiplied by 0.2 60. Wall area (700 sq. ft.) multiplied by 0-025 17.; 227.5 Substituting in (i) above, we now have: 0.05 (3000) t = • = 0.66 seconds This is nearer a satisfactory value, and may give good results on the air. If, now, one of the 15 by 10 walls of the room, preferably that nearest the contemplated position of the microphone, be covered with curtains having an absorption coefficient of 0.5, the total absorption of the studio becomes 298.7, a value which the reader may readily verify by going through the simple computation himself. The reverberation time is then reduced to o. 5 second. In order to provide margin against reflecting surfaces of the piano, tables, etc., and to allow variation of the period to secure different effects, it might be well to drape two of the walls of the room with curtain material suspended from rods and movable at will, and to have the rug only partly covering the floor. This will permit variation of the reverberation period between, say, 0.3 and 12 seconds, which gives sufficient latitude for experiments. Various other combinations are of course possible, and each designer may work out schemes to suit his own requirements or fancies. In selecting patented sound proofing materials it is well to try to secure a curve showing varia- tion of the absorption coefficient with frequency. The nearer this characteristic approaches to the ideal flatness between 100 and 5000 cycles, the more favorably it should be considered for gen- eral use. It is better to choose material with a moderate absorption coefficient which remains fairly constant over the audio range, than a highly selective surface which has high absorp- tion at one or two pitches and reflects badly at other points. Formula (i) above is not as accurate as a later equation worked out by P. E. Sabine, given by Crandall in his Theory of Vibrating Systems and Sound (D. Van Nostrand Co.): t = 0.0083 V (9-1— IOKIO A) A (2) Formula (2) gives results for I slightly lower than (i), so thdt for electrical reproduction it is just as well to use the simpler expression (i), which gives additional margin against excessive reverberation. Where lively surface materials are used in studio or auditorium design, a protection against echo and objectionable reverberation is to break up flat surfaces by recessing, coffering, etc., thus diffusing the reflected energy. The circular of the Bureau of Standards re- ferred to at the beginning of this article gives a table of allowable number of instruments in a room of given volume, for best artistic perform- ance. This data should interest musical directors and broadcast technicians who habitually overcrowd their studios: VOLUME OF ROOM 50,000 100,000 200,000 500,000 NUMBER OF INSTRUMENTS 10 20 30 60 While allowing 5000 cubic feet and over per instrument may be excessive, this table is prob- ably nearer a sensible compromise than the comical broadcast practice of jamming thirty