International photographer (Jan-Dec 1934)

Eighteen T h INTERNATIONAL PHOTOGRAPHER July, 1934 SOUND PROOF STAGE DESIGNING By Paul S. Harmer Sound stages are no longer the nightmare to the motion-picture producers that they were in the early days. Much has been learned. Some things had to be changed as improvements were made in sound recording equipment. When sound first came to the studios of Hollywood, Dr. Knudson, of the University of California, Los Angeles, fortunately was equipped to carry on experiments in his sound testing laboratory at Westwood. He found definite figures for transmission and reflection of sound with the various types of walls which were tried. Had it not been for his co-operation right here at home the problem would have been more costly and haphazard. Much of the early efforts in the sound proofing of stages were made by men who had worked in radio broadcasting stations ; the principles involved were similar, but a suitable motion picture stage has been found to be an entirely different construction problem. The major studios have found that a stage 80 ft. wide and 150 ft. long by 30 ft. clear under the trusses; or 110 ft. wide by 200 ft. long and 40 ft. clear under the trusses will suit almost any condition that may arise. A single unit stage is by far the most efficient. Some studios have built two and three stages adjoining each other, all under one roof. The studio saved the price of an outside wall, but they have paid for it many times since in delays and spoiled sound track, caused by noise from the adjoining stage, this noise following the floor joists or passing through the wall and doors. Steel framework and steel trusses have always proven superior to wood trusses, especially since earthquakes must be considered when designing any new building. Wood studs, joists, rafters, sheeting and flooring are better non-conductors of sound than steel ; consequently they are preferable because of this quality. No resonance trouble is encountered from any ordinary building material, the fundamental frequencies are usually below 50 cycles ; for instance, pine wood measured 7 cycles per second; iron measured 12 cycles; plastered concrete measured 28 and plastered brick measured 48 cycles. These are relative values and will vary according to mass and volume. The partial modes give rise to overtones two and three cycles higher, but these are too weak to cause any trouble. Strange as it may seem, frequencies below 500 cycles carry through building walls and reverberate longer than the others; frequencies from 500 to 4,000 cycles are intermediate, while the frequencies above 4,000 are quickly absorbed and fade out. This is the reason that calculations on the low frequencies are so important when designing a new stage or sound proofing an old one. This also is the reason the hard wall sets have come back into use; in order to preserve the higher frequencies in more equitable proportion. In order to make clear the intensity of sound as expressed in decibels, I will illustrate the difference between these intensities as expressed in numbers. A bel is the intensity of sound in dynes per square centimeter for a given frequency, and a decibel is one-tenth of a bel, or ten decibels equal one bel. The following is a logarithmic table. Numbers expressed in exponents. And so on up to 130 decibels, where sound is so intense it becomes painful. A sound of 70 decibels level is 1,000,000 times more intense than a sound of 10 decibels level. It is practical to build a sound stage which will have a sound transmission loss of 60 decibels at 128 cycles, providing the neighborhood noise warrants such a building. Fig. I shows a typical wall and roof of such a structure. This is one of the first large stages built for M-G-M and is located betwen the Pacific Electric Railroad and Washington Boulevard. Fig. 4 shows a typical wall and roof of a recently built stage in the same studio ; instead of reinforced concrete they used steel trusses and steel frame ; wood studs, joists and ratfers doing away with the concrete floor, this is typical of the advances made during the past six years. It is an expensive proposition to build a large stage with a 60 decibel transmission loss and a reverberation period as shown in Fig. 8. Consequently, quieter locations are sought, or the building is constructed to a less expensive specification. A stage with a 50 decibel transmission loss at 128 cycles can be built either with a double or single wall. The double wall type has proven to be the best. Roofs and ceilings generally do not give any trouble, even when airplanes fly overhead at a reasonable distance ; trucks, tractors, hammering and loud talking are the despair of the sound mixer. Extraneous noise which exceeds 6 decibels in the set, where a company is recording dialogue, is sure to cause a retake, and while some shots can stand more than 30 decibels interference it is not recommended. To illustrate the sound insulation value of various walls, a 2 by 6 studded wall with metal lath and 1 inch of gunite plaster has a transmission loss of 30.6 decibels at 128 cycles; the same wall with l/2 inch plasterboard added to the inside, has a transmission loss of 34.4 decibels. The same wall, except using % inch fibre board instead of plasterboard, has a transmission loss of 37.4 decibels. There is a difference in the kind of plaster. For instance, a 2 by 6 studded wall with metal lath and two coats of gypsum plaster has a transmission loss of 40 decibels at 128 cycles. It is an easy matter to construct a sound stage with a transmission loss of 45 decibels, and a reverberation period of \y2 seconds, at 128 cycles, but as you go into walls of greater transmission loss and shorter reverberation periods, the costs mount quickly to the benefits accomplished. Fig. 2 is a typical wall at Universal ; Fig. 3 at the small studio on Beachwood ; Fig. 5 at Darmours ; Fig. 6 at United Artists and Fig. 7 at RKO-Pathe. Fox Hills stages correspond to Fig. 1, while their Western Avenue stages have not been treated for sound, except by hanging large ozite blankets, 1 inch thick, about a foot 10" = 1 1111 = 10 lir = 100 Hi = 1,000 10* = 10,000 106 = 100,000 10" = 1,000,000 107 = 10,000,000 then logi, 1 = 0 10 = 1 100 = 2 1,000 = 3 10,000 = 4 100,000 = 5 1,000,000 = 6 10,000,000 = 7 — 0 bels — 0 decibel* — threshold of audibility. 10 20 30 40 50 60 70 above Please mention The International Photographer when corresponding with advertisers.