History of Sound Motion Pictures by Edward W Kellogg Third Installment

Home , Fantasound, Optical sound, Phototube, RCA

History of Sound Motion Pictures

by Edward W Kellogg

Third Installment History of Sound Motion Pictures by Edward W Kellogg

Our thanks to Tom Fine for finding and scanning the Kellogg paper, which we present here as a “searchable image”.

John G. (Jay) McKnight, Chair AES Historical Committee 2003 Dec 04

This material is posted here with permission of the SMPTE. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the SMPTE.

By choosing to view this document, you agree to all provisions of the copyright laws protecting it. FINAL I NSTA-LLMENT

History of Sound Motion Pictures By EDWARD w. KELLOGG

For the abstract of this paper which was presented on May 5,1954, at the Society's ing is reported by J. Crabtree in the Convention at Washington, D.C., see the first installment published in the June October 1933 and February 1934 Jour- Journal. The second installment appeared in the July Journal. See p. 437 for nals.232 errata in second installment. In 1934 C. N. Batsel described a nonslip contact printer2= employing a principle proposed and demonstrated earlier Developments Which Extended differs from this assumed average. By me- by A. V. Bedford for a different applica- the Frequency Range chanically stretching whichever film is tion.7@,234Bending a film stretches one shorter, two films can be moved together face and compresses the other, and it is Better Light-Modulating System for Vari- through an appreciable distance in per- only necessary that the contacting sur- able Area. Mention was made in the fect nonslipping contact. A printing sys- faces of the two films be made equal. section on loudspeakers of a demonstra- tem in which this was done was developed In the nonslip printer, the negative is tion by Dimmick and BelaP9 of sound by the Technicolor engineers for their rolled through the machine at fixed with extended frequency range. Aside color transfer, but it was not used for curvature, and the print stock (held from the 'improved loudspeakers and the sound prints. An extended investigation against the negative at the printing ribbon microphone (whose response was of the losses and irregularities in high-fre- point, where it is propelled solely by practically uniform from 60 to 10,000 quency response which result from slip- friction) is made automatically to assume cycles) there were a number of advances page and imperfect contact during print- the curvature at which identical numbers that contributed to the result. The new galvanometer and optical system made so much more light available that it was feasible to reduce the width of the April 8, 19.30. A V BEOFORD 1,754,187 recording beam to 4 mil, thus improving 8ELT OR STRIP DRIVING ARRANGEMENT resolution. No small factor in giving clean high frequencies is avoidance of flutter, Filed Nov 5. 1927 particularly rapid flutter such as 96 cycles. In the demonstrations, both Fig. I recording and reproduction were on magnetic-drive machines. * (The rotary R stabilizer was not yet available.) E f-' Ground-nois: reduction was by galvanometer bias with a single narrow line of transparent film when the modulation was zero, and it is my recollection that a measurement indicated a ratio better than 50 db between signal at full modulation and the ground noise when biased for zero modulation.

Nonslip Printer and Improved Sprocket- Type Printers. With the sprocket-type contact printers in almost universal use, a certain amount of slipping of the negative with respect to the print film is almost inevitable. The curvature at the sprocket compensates for a certain negative shrink- Fis.2. age at which the ratio of radii of the shrunk negative and unshrunk print stock is just equal to the ratio of their lengths. The sprocket diameter is designed to make this compensation correct for an average negative shrinkage, but it will be only ap- proximate for negatives whose shrinkage

*The special film-phonograph used in the demonstrations was a prototype of those used in the Disney Fantasia reproduction (see Fig. 7, p. 136, Jour. SMPE, Aug. 1941). The film was pulled by the magnetically driven drum over a curved supporting plate where the tracks were scanned, and was steadied at the other end of the plate by another drum with flywheel. This was an anticipation of the double flywheel system Fig. 17. Control of film speed .by flexure (A. V. Bedford). Principle later used in now used with magnetic tape and film. nonslip printer.

422 August 1955 Journal of the SMPTE Volume 64 of sprocket holes are fed through in a comes up only with 35mm-to-35mm nition was the fact that ultraviolet light given time. printing, for which optical printers are is more rapidly absorbed in the emulsion The nonslip printer was not built for rarely used. than is visible light. The increased sale by RCA, but free license and draw- In spi%e of such improvements as absorption must be compensated for by ings were offered, and numerous labora- have been made in developer turbulation increased intensity, but the net effect is tories built and used them, notably in the processing machines, the sprocket that the exposure is confined more nearly Consolidated Film Industries in Holly- type of printer does not appear to have to the surface where the silver halide wood and Ace Film Laboratory in any strong competitor for density prints. grains are more completely used, and the Brooklyn.23 In such a situation the efforts of several result is less graininess. Most important A printer based on the identical prin- Eastman engineers”2.113 to improve of all is that the light scattered by the ciple was developed independently by sprocket action are well justified. The emulsion does not spread as far sidewise, R. V. Wood.”6 shrinkage-compensating sprocket de- and thus enlarge the image, as does Although nonslip printers were not scribed above under “Mechanical Sys- visible light.N2s2a made in large numbers, and did not to tems” should result in great improve- Dimmick reported the results of tests any great extent displace the con- ments. The new film-base materials with with ultraviolet in August 1936.2u He ventional sprocket type of contact printer, their reduced shrinkage should also make had found it possible to obtain adequate they served to demonstrate how much for better results with printers of the exposure with the regular incandescent improvement was possible through better sprocket type.391 lamps and Corning No. 584 filters. printing, and in this way stimulated Conversion for ultraviolet recording makers and users of sprocket printers to Optical Printers. Cost and quick produc- involved provision for the filter, redesign improve their machines and maintain tion have been dominating considera- of the objective lens, and substitution of them in the best possible condition. tions in much 16mm production, and glasses with less ultraviolet absorption In a number of laboratories it appeared contact printing on the sprocket has pre- for other lenses in the system. Many of that better prints resulted if the teeth vailed, the printers being simpler than the variable-area recording systems were were removed from the side of the other types and available in large num- converted. Single-film newsreel systems sprocket next to the soundtrack, leaving bers to meet the heavy wartime de- (with the sound recorded on panchro- only a smooth supporting rim. The mands. However, registration of nega- matic negatives) were especially benefited theory behind this is not clear unless it tive and positive perforations is not a by use of ~ltraviolet.~~~,~~~ is that such disturbances of contact as factor in the quality of 16mm prints of The results obtainable by the applica- are due to film sticking on the teeth are either density or area tracks. There is tion of ultraviolet exposure to variable- thus kept further from the soundtrack. thus no obstacle on this score either to density films were studied and reported There were two reasons which limited nonslip printers or to optical printers by Frayne and Pagliarulo in June 1949.247 the use of the nonslip printer. Since with independently filtered film motion, They found major improvement in there is a very narrow region of close and both types have been used, the definition and reduced distortion, but contact: the printing beam had to be nonslip for printing from 16mm nega- only about 1 db reduction in ground narrow. Under such conditions any tives, and the optical principally for noise. Further experiences with ultra- speed irregularities cause variation in printing on 16mm film from 35mm nega- violet are reported by Daily and print exposure, which sometimes be- tive~,=~-~~~but also with a 1 : 1 optical Chambers.248 comes audible in density prints. While system for printing from 16mm negatives Many printers were modified to print there is no valid excuse for such speed (Precision Film Laboratories, New sound with ultraviolet. High-intensity variations, this trouble contributed to the York) mercury arc lamps were widely used for conclusion that the nonslip printer was The Eastman Co. has designed several the printers. unsatisfactory for density printing. Film models of optical reduction printers Ultraviolet exposure gives lower laboratory operators were naturally in which the 35mm and 16mm films gamma for the same development than averse to maintaining one type of printer are on sprockets on opposite ends of a does white light. Advantage has been for area and another for density. A single shaft and a U-shaped optical taken of this in certain cases of density more fundamental fault of the nonslip system is employed.241This system mini- recording to reduce the gamma of the printer in making density prints is due mizes chances of slow “wows,” but de- sound print without departing from to the fact that negative and positive pends for good results on the degree of optimum development for the picture. perforations are not maintained in excellence attainable with sprocket action. register. The more active developer The shrinkage-compensating, radial- Coated Lenses. Each time any optical circulation close to sprocket holes tends tooth sprocket described in the section or light-source improvement made an to darken these areas, resulting in a on “Mechanical Systems”I13 was de- increase in image intensity possible, the slight 96-cycle hum. The effects on the veloped with printer applications par- benefit could be realized in terms of negative and print are compensatory ticularly in view. finer resolution and better high-fre- provided the printing is done with the quency response. holes in registration, but cumulative if Ultraviolet Light for Recording When the results were published of out of register. With a purely friction and Printing some tests in which the reflection from drive at the printing point the relative For truly distortionless recording the glass surfaces had been materially re- positions of positive and negative per- ideal light spot would be of infinitesimal duced by applying quarter-wavelength forations are unpredictable and may width. In practice, as various improve- coatings of certain low-index minerals,24g drift slowly between one condition and ments made light of greater intensity Dimmick procured equipment, mastered the other. This results in a hum that available, the nominal width of the the techniques of applying such coatings comes and goes and is therefore more recording spot was reduced (to 0.00025 by evaporation in vacuum, and com- conspicuous than if steady. Area tracks in.), but a limit is set by diffraction. pared the merits of various materials are far less sensitive to such develop- With light of shorter wavelength lenses and methods of hardening the deposited ment variations. of the same dimensions can form a more layer. Having found a satisfactory pro- Optical printers with independent nearly perfect image. This is one of the cedure, he treated all the glass-air filtering of negative and positive film considerations that led to a study by surfaces in the RCA recording optical motion would be subject to the same Dimmick of the possibilities of recording system. Although the loss at each glass- uncertainty with respect to relative with ultraviolet light. Even more im- air surface is only about 5%, there are perforation positions, but this problem portant than the improved image defi- sixteen such surfaces (excluding the

Kellogg: History of Sound Motion Pictures 423 lamp bulb), and by reducing each to a Measures for Reduction and Control errors) of recording or reproducing slits magnitude of between 1 and 20/,, a gain of Distortion in the case of area recordings, have been of more than 50% was possible in the There is hardly space here to mention published by Cook260and by Foster.261 brightness of the image on the film. the many refinements, including prin- In the case of density tracks the effect Not only is the image thus brightened, ciples of good design, by which distortion of misalignment is practically equivalent but the stray light due to “lens flare” has been kept low in microphones, ampli- to widening the scanning slit, the results is reduced to a small fraction of its fiers, recording devices, optical systems of which in terms of high-frequency loss previous magnitude. A review of the and reproducing systems. We shall con- are given in a paper by Stryker.26z subject, containing further references, fine ourselves to discussing the sources Both Cook and Foster have published has been published by W. P. Strick- of distortion associated with photographic analyses of the distortion in area record- land.250 soundtracks. These may be listed as: ing due to slit width.2@s2@ The first RCA coated-lens optical Calculations of the distortion resulting system to be field-tested was a “variable- (1) background noise caused by graini- from certain cases of uneven slit illumina- intensity” system taken to Hollywood by ness (especially in density tracks) or by tion and with several types of area track Dimmick early in 1938 and used experi- scratches or dirt in the track area (most have been published by Cartwright and mentally during some months at the serious in area tracks) ; Bat~el.~~~A test film has been made Fox studios, C. N. Batsel being the (2) noise (hum) and amplitude modu- available for checking light uniformity.266 RCA liaison engineer.251I believe that lation resulting from sprocket hole It has a number of very narrow record- I am right in saying that the demonstra- proximity; ings of a tone in a series of positions tions of this system were partly responsi- (3) fluttter or ‘‘wows;’’ across the track, and their relative ble for arousing interest in the benefits (4) high-frequency loss; outputs when played indicate light to be had by such treatment of glass (5) irregular high-frequency loss, and intensity at the corresponding positions surfaces. modulation of recorded sound, due to in the scanning beam. Within a short time the lenses in a printer defects; number of the RCA recording systems (6) waveform distortion due to mis- Sprocket Holes and Irregular Development. were given low-reflection coatings, and alignment of the aperture (slit) either The proximity of sprocket holes to the the practice of coating lenses in both in recording or reproducing, and to soundtrack causes various difficulties. area and density recording systems uneven reproducing-slit illumination ; Mechanically the variations in stiffness spread rapidly.252 (7) waveform distortion and produc- of the film cause it to bend in a form tion of spurious tones due to nonlinear resembling a polygon instead of a true Dichroic Rejectors. As a by-product of input-output relations (more frequent in cir~le.~~~,~~’This causes 96-cycle flutter, the work on low-reflection coatings, density recording than in area) ; less perfect contact in printers, and 96- Dimmick studied multilayer coatings of (8) waveform distortion and produc- cycle modulation of high-frequency tones alternate high- and low-index ma- tion of spurious sounds due to finite if optical systems have small depth of terials.2” With these he was able to width of the recording slit and image focus and are not exactly focused for the produce plates having very high re- spread in the film emulsion (a problem average emulsion plane. More serious flectivity in one part of the visible in the area system) ; and from the sound standpoint is the effect spectrum and very high transmission in (9) distortion due to overloading. of the holes on devel~pment.~~~.~~~ another (65% reflection at 7500 A and In film development there is always a better than 95% transmission between Measures for improving sound records tendency to nonuniform developer activ- 4000 A and 5000 A). with respect to background noise are ity due to local exhaustion. For ex- The first motion-picture applicatioq discussed in part under “Ground-Noise ample, over and near a large dense of his “dichroic reflector” was for re- Reduction,” but also include extreme area the developer is slightly weaker in its flecting the red light of the recording care as to cleanliness (especially in proc- action than over the middle of a lightly light beam to a caesium phototube for essing), to hardening and lubricating exposed area, and this in turn is weaker monitoring, while losing practically none (waxing) the film,259to maintenance of than the average of the bath. If the of the blue and violet light that pro- projector condition, and to the several partly exhausted developer as it diffuses duces the exposure.2M expedients by which the maximum re- out of the emulsion is not immediately During the war there was heavy produced volume can be increased. carried away from the film by currents demand for dichroic reflectors, designed Fine-grain film has made perhaps the in the liquid, it may creep along the to various specifications, for the armed greatest single contribution to improve- film surface and weaken the develop- forces, and also great demand for low- ment in ground-noise ratio, especially ment in such areas. If there is little fluid reflection treatment of many optical for density recordings, but its benefits movement except that caused by the devices. Selective reflectors of this type have been in so many directions that travel of the film, the direction of travel have also played an important part in only the present topic heading (Reduc- is often evident from the appearance of color-television devel~prnents.~~~ tion and Control of Distortion) seems pict~res.~~~~~~~Circulation of the liquid broad enough to include it. * is somewhat freer near the holes and Heat- Transmitting Rejectors. Another Various developments which improved edges of a film than elsewhere, and outcome of the development work on high-frequency response have already therefore the development and average dichroic reflectors is a design in which been mentioned, and to these should of density greater. This was discussed in the transition from high to low reflectivity course be added the use of fine-grain connection with printers. occurs just beyond the end of the visible film. The topic “Improved Printers” The irregular development troubles spectrum, so that as a mirror it would has also been included in the section on are reduced by very active stirring or compete well in efficiency and whiteness better high-frequency response. Reduc- turbulation of the developer while the with the best silver or aluminum mirrors, tion of flutter is dealt with in the film is passing thro~gh.~~2Improvements while the invisible heat rays are better section on “Mechanical Systems.” on this score are reported by Leshing, than 75% tran~mitted.~~~~~~~No compar- Analytical studies of the distortion that Ingman and Pier.273 ably effective means of separating the results from misalignment (or azimuth Uniformity of development is further useful visible from the infrared radiation improved if the emulsion has such has been hitherto available, and we may * It has been said that the entire evolution of sound reproduction can be divided into two parts: characteristics that its “gamma infinity” look forward to extremely valuable (1)~helearning how to make a noise and (2) the is only slightly higher than the desired application^.^^^^^^^ reduction of distortion. gamma. Film manufacturers have been

424 August 1955 Journal of the SMPTE Volume 64 successful in producing fine-grain emul- image spread increases with exposure. In the earlier years of recording, the sions which come much closer to this The distortion is largely in the form of a conditions for linear relation between desideratum than earlier types. change in average transmission when a negative exposure and print transmission high-frequency wave is present. It is were figured out by conventional sensito- Waveform Distortions. The sensitometric often referred to as “zero shift.” It was metric methods, with correction for the studies and analyses about which we particularly troublesome in some of the ratio of projection (or semispecular) have already spoken are a part of the earlier 16mm recordings. It is possible density to diffuse density. But this is large body of literature bearing on the however, by selecting the proper print- tedious, and a testing system which subject of distortion of the nonlinear ing exposure, to make the image spread duplicated soundtrack recording and type in density soundtracks. There have in the print neutralize that in the reproducing conditions was found prefer- also been a number of experimental negative .m6 able. In 1935 F. G. AlbinZi9described a studies reported in this Journal giving In January 1938, J. 0. Baker and dynamic check on processing which distortion measurements with both den- D. H. Robinson described tests and consisted in making several recordings of sity and area tracks. Two of the earliest equipment which became the standard a tone, all at the same fairly low ampli- of these were by Sandvik and Hall method for ascertaining the correct tude, with a series of increasing mean (October 1932)274 and by Sandvik, relative negative and print den~ities.’?~ valve-ribbon spacings. Comparison of Hall and S treiffert, October 1933.275 For 35mm tests a 9000-cycle tone is the relative outputs from the print One of the most effective measures for 100% modulated at 400 cycles, but showed whether the slope of the curve reducing such distortions (second in there must be no 400-cycle component of print transmission vs. negative ex- importance only to avoiding overload) in current supplied to the galvanometer. posure was constant or not. This was is by the use of push-pull soundtracks, For a given negative exposure and something of a forerunner of the inter- and this applies about equally to density development, a series of prints having modulation test. In the latter a tone of and area systems, although the causes different black area densities are made. 1000 cycles or higher is superimposed of their distortions (Nos. 7 and 8 in the They are then reproduced and the on a low-frequency tone of, say, 60 list) are quite different. magnitude of the 400-cycle output cycles. Normally the level of the high- Two systems of testing for distortion measured and plotted (in db below the frequency tone is 12 db below that of the (one for area and one for density systems) 9000-cycle level, or other reference) as a low tone, and the combined amplitude have assumed great importance since function of print density. The curve goes just short of overload. The print is run their introduction for ascertaining opti- through a sharp minimum at optimum through a reproducing system, and the mum ranges of exposure of negatives print density. With a negative of greater reproduced high-frequency tone sepa- and positives. These are known as the (black area) density the optimum print rated out and measured for fluctuations in “cross-modulation”17~ and the “inter- density is found to be higher. One of the amplitude. Nonlinearity (distortion) be- modulation”277tests, and will be briefly most frequent problems is to select a comes evident as fluctuations of the high- described. recording exposure which will give frequency tone, in general at rates equal maximum (or at least satisfactory) to the frequency of the low tone or cancellation of cross modulation, when multiples of that frequency. Overload. Overload in any element printed for a certain desired print of the entire sound system other than in density (say 1.4 to l.6).* Fine-Grain Films.280-282While the many the soundtrack (either in the original Experience shows that if the 400-cycle components of sound recording and recording or in the re-recording) means cross-modulation is 30 db or more below reproducing systems were being im- that the system has not been properly full modulation, the sound is satis- proved, the great film companies were designed. It is easy to say that the way factory. In similar manner if the print busy. In the period 1930-1940, the to avoid distortion due to overloads is exposure and development are specified, Eastman Co. offered at least seven new not to overload, but that is asking too the optimum negative density can be emulsions,283each of which offered some much. The price in terms of reduced determined. For 16mm processing con- advantages for soundtrack application average level is from a practical stand- trol, a 4000-cycle tone is modulated at as compared with the currently used point excessive. The levels set in record- 400 cycles. In push-pull recording the picture positive (Eastman No. 1301). ing are a compromise between too low an cross-modulation in the two parts of the In 1932 Eastman brought out No. 1359, average level (with ground noise there- track largely cancel each other and intended for variable-density recording fore more conspicuous) and too many tolerances are extremely broad. It was however similar to No. 1301 and too bad-sounding overloads. * in contrast. In 1936 No. 1357 was In sections to follow, a number of ex- At the 1954 spring convention Singer and McKieZi8reported good results with brought out for variable-area recording, pedients are mentioned for compassing a slightly faster than No. 1301 without wider range between background noise an electrical compensating circuit, so that cross-modulation can be practically sacrifice with respect to grain and resolu- and maximum reproduced level. These tion. With ultraviolet exposure it proved devices may be used either to make the neutralized when making direct positives. This involves determining what popular for variable-density recording. background quieter or the loudest noises For release prints, however, the positive louder, and if properly applied can ac- function of frequency, exposure and (possibly) amplitude comes nearest to No. 1301 continued the almost universal complish the latter with reduced instances choice except as similar films of other of soundtrack overload. expressing the magnitude of the cross- modulation products. manufacture were used. What was primarily wanted was finer Cross-Modulation Test to Determine Best Intermodulation Test for Variable-Density grain, for this would mean cleaner, Processing for Variable Area. While the Control.n7 In the June 1939 Journal sharper images for variable-area tracks, variable-area system is not critical to Frayne and Scoville describe a test for less background noise in variable-density gamma product it is subject to distortion, control of variable-density processing. tracks, and better resolution and high- particularly at high frequency, due to frequency response in all tracks. But the spread of the exposure outside the *This test is made simpler than it might appear by fine grain in general means lower speed, area of the actual image, caused by the the fact that when a high-contrast negative is or more exposure required, and too great used, the print black-area density is scarcely sacrifice in this direction would make scattering of light in the emulsion. This affected by differences in the black-area density of the negative, hence the printer light setting new films unacceptable. Another diffi- *Most movie fans have probably forgotten what can generally be kept the same for the entire culty was that with many of the earlier a gun sounds like. series of negatives. fine-grain emulsions the image had a

Kellogg: History of Sound Motion Pictures 425 brownish tone, which was objectionable these reports indicate that while a re- No. 1302 as compared with 2.1 for No. in the picture, and so ruled these emul- recording print might be 8 to 12 db better 1301.290 sions out for the release prints. Measures in signal-to-noise ratio than one made were found for correcting this fault in on the previous types of film, only 4-to-6 Variable-Density vs. Variable-Area.291J92 Eastman fine-grain positive No. 1302. db net gain was to be expected in the The rivalry between advocates of vari- The problem of exposure was less serious release print if this were on the usual able-density and variable-area recording in the case of fine-grain films for release stock. This indicates why the real gains is nearly as old as commercial talking and for duplicating than for recording from use of fine-grain film were not pictures and has unquestionably pro- stock, although some changes in printers achieved until they were used for the moted progress in both systems. The were required. Special films were intro- release prints as well as in the earlier fact that the industry did not standardize duced about 1937 for duplicating pur- stages. (See also references 287-290.) on one or the other track does not seem pose~?~~ Some difficulty was encountered in to have been any handicap, for all One of the chief problems in variable- processing the negatives to a suitably theaters could play either type of track. area recording was due to image spread low gamma, but it was accomplished by During the earlier years of commercial (beyond the exposed area), but it was use of special developers. (Uniformity of sound, the advantage seemed to be on found possible to make image spread in development is almost impossible if the the side of area for music but density the print compensate for that in the time is cut so short that the gamma is for speech intelligibility. With both at negative. The criticism of one of the far below the “gamma-infinity” for the their best there was little to choose in trial emulsions intended for variable- given emulsion and developer.) In clarity of speech reproduction, but the area negatives (No. 1360) was that it order to permit somewhat higher values density system seemed able to take more was too good in this respect for the of gammas in the fine-grain negatives the abuse without too serious loss of articula- currently used print stock. practice was adopted in some labora- tion. Before control of zero-shift was Re-recording is the universal practice. tories of holding down that of the fine- well in hand (see cross-modulation Graininess and other imperfections in grain release prints by printing the test276)this type of distortion seemed to the original negative are transferred to sound with ultraviolet light, which gives do more damage to the quality of the the print, passed on to the second nega- lower gamma for a given development high-frequency reproduction than the tive by the re-recording operation, and than does white light. nonlinearity to which the density system in the final printing are added to the In 1938 the Eastman Co. brought out a is more subject. With improved tech- graininess of the release print. The use of high-resolution film (No. 1360) for niques and equipment this difference direct positives offered promise of elimi- variable-area recording and in 1939 a disappeared. nating at least one step which was con- recording film (No. 1366) for variable Another factor which for a time was tributing its share of noise. In 1937 density. In 1941 the No. 1302 fine- prejudicial to the area system was that, Eastman introduced its No. 1360, and grain positive which has been widely as compared with density recordings, Du Pont its No. 216, both fine-grain used for release prints, became available. low-level passages seemed to be ex- films designed for making variable-area In 1945 the Eastman Co. brought out a cessively weak when the controls were direct positives. high-contrast fine-grain film (No. 1372) set for satisfactory normal and high-level Changing to a new release stock repre- for variable-area recording and a low- passages. One theory was that, due to sented a greater hurdle than the use of contrast fine-grain film (No. 1373) some fault in equipment or system, there special films for negatives and for the for variable-density recording, which was “volume expansion,” or actual re-recording print, so the general thought could be processed in regular picture- exaggeration of level difference. To was to evolve the best possible films for negative developers instead of requiring those most familiar with area recording these applications, and take what im- special low-energy developers. The same such a theory appeared untenable. provement could be gained in this way. year Du Pont brought out a new fine- Measurements did not bear it out. As the story turned out, use of the fine- grain recording negative (No. 236) for Monaural or single-channel listening grain films did not become general until variable density, whose contrast could be itself makes level differences seem they were used also for the final release more easily controlled than previous greater than direct binaural listening. prints. types because of low gamma infinity. Could it not be that the density record- Film graininess contributes to back- The fine-grain recording films are all ings by their nonlinearity produced a ground noise in much greater measure in much slower than the previous recording compression effect? I recall that one of the density system than in the area films, (No. 1302, for example, requiring the most confident exponents of this system, but the requirements seem to about four times the exposure of No. explanation was M. C. Batsel. One of have been harder to meet. Many tests 1301), and recording optical systems, the fundamental differences in the were run and studies made during the although improved by coated lenses, systems is that an area track overloads years 1938-1940, and a special com- could not with tungsten lamps provide very abruptly, whereas in a density mittee representing those particularly sufficient exposures for the combination track the upper and lower limits of interested in variable-density recording of fine-grain and ultraviolet light, but light transmission are approached more reported in the January 1940 use of the fine-grain films with white- gradually. With such a characteristic, in which tentative specifications were light exposure went so far in affording considerable overload (i.e. beyond the given to guide in the development of the benefits of high resolution and low range of true linearity) can be tolerated the needed fine-grain films. noise that the combination seemed without too objectionable effect, and it Du Pont No. 222 was one of the first hardly needed. On the other hand, was thought that considerable advantage fine-grain films to gain considerable printing fine-grain positives with ultra- had been taken of what might be called acceptance, especially at the M-G-M violet light from high-intensity mercury permissible overloading in the density and Paramount studios, where it was arcs has been widely employed. recordings as compared with those made used for original negatives, re-recording It is of interest that the Callier co- on the area system. prints, re-recorded negatives and for a efficient (ratio of specular to diffuse A nearly equivalent effect in an area limited number of release prints. High- density) is less with fine-grain than with recording can be obtained by rounding pressure mercury lamps were substituted coarser-grain films, and the result is that the corners of the V-shaped opening in where needed to give the required print- a higher control gamma (in which the the aperture plate (whose image is the ing exposure. Experience at the Para- measurements are of diffuse density) light spot on the slit as shown in Fig. mount studios is reported by C. R. Daily is needed to give the optimum picture 13). Transition curves are introduced at in the same issue.286Figures published in on the screen-for example 2.5 for the outer (full-slit illumination) ends,

426 August 1955 Journal of the SMPTE Volume 64 while the vertex is drawn out to a finer optical syatem by changing aper- compressors at the Warner Bros. Studios point. Such curved apertures were tUres.183,184,297 is reported by B. F. Miller.301 tried and largely overcame the cause of The strict adherence to one system criticism. (area or density) on the part of the major Squeeze Track.307In view of the limited However, the use of electronic com- picture companies has in considerable range of loudness which a soundtrack pressors (long employed in broadcasting measure broken down, and several are permits, and the very great range en- studios) appeared preferable to sacri- employing both systems, depending on countered in our ordinary experience, ficing the range of low-distortion light the type of operation required.298 For many expedients were tried for the modulation. Compressors were intro- example M-G-M, while continuing to purpose of increasing the range obtain- duced in the recording channels with use double-width push-pull density for able from the film. One of these was the very satisfactory results. The first such initial recordings,* releases on variable “squeeze track” described by Wesley C. compressors were tried at RKO and area. One of the reasons for changing to Miller of M-G-M.307If a variable-density Warner studios and compression became area for release prints is that projection- track is reduced in width, both the noise the standard. If the original recording is ists often neglected to change their gain and the modulation are reduced. The on a wide-range (or low-background- settings when switching back and forth ground-noise reduction system described noise) system, most of the compression between area and density tracks, with a by Silent and Frayne in May 1932lS8 is introduced in the re-recording to the result to the disadvantage of the film may not have been as yet available, but release-print negative, but I am told with lower output (note also interest in the two devices are notequivalent and can that even with magnetic recording some Perspecta Sound.) be complementary. There is a practical compression is used (or at least avail- Mention is made, in the discussion of limit set by film characteristics as to able) in the initial recording. control tones, of a system in which the how far the ground-noise reduction, Any simple and general statement of control tone is of subaudible frequency by reducing mean negative exposure, the relative signal-to-noise ratios of the and superimposed on the recorded sound. could be carried, and this was at about two systems is impossible, since each Modulation of the sound by the control 10 db of reduction, but the noise can be has its own cause and type of ground tone is much more easily avoided in still further reduced by narrowing the noise: grain hiss in density, and ticks area recording. track. The controls in recording or in and pops due to dirt or scratches in the re-recording are used to avoid too low area system. Area starts with an initial Compressors.299In the previous section levels on the track, but some of the range advantage of about 6 db greater output, comparing the area and density systems, can be restored by narrowing the track and if the film remains in good condition the first urgent need of electronic com- in the release print. There were several its signal-to-noise ratio is better. The pressors was described, but their use has ways in which this could be done; variable-area system was chosen by the proved so advantageous as to have be- one, for example, was by preparing a Bell Laboratories engineers for their come quite general, both in recording masking film and running the print stereophonic demonstrations. and re-recording channels.300Electronic through the printer a second time with As various improvements became compressors had long been in use in radio the track width determined by the mask. available to the industry they worked to broadca~ting.~~~~~~~~In their application the benefit now chiefly of one system to sound recording, if there is no pro- “Reversed Bias” System.3o8An expedient and now of the other. Fine-grain films vision for re-expansion, they do not for obtaining greater output for bursts of helped both, but helped density more. actually increase the reproduced volume high level sound was described by Hansen and Fa~lkner~~~of Twentieth The higher output from variable-area range although they may seem to. A tracks led to the proposal by LevinsonZg3 common characteristic compresses 20 Century-Fox. In effect the light-valve that intercutting of the two types of db into 10 db at a uniform rate. This bias operates in the usual way for normal and low levels, but for passages of extra track be used as means for increasing would rob any passage of expression reproduced volume range. only to a slight degree. As the overload high level the mean ribbon spacing is increased to as much as twice normal The RCA light-modulation system point is approached it is common to make the compression more drastic, while still being fully modulated. There has been modified to record variable- is some resulting distortion, but sub- density, by use of an out-of-focus image for example 20 into 3 db. This is sometimes called a limiter type of compressor. stantial increase in print output. The or “penumbra” on the slit in place of If (for example by means of control loss of some relatively high-frequency the usual sharply focused triangular output is probably not objectionable. spot.183r294This system modulated the tone) provision is made for re-expansion, light reaching the film by changing its the compression may be such as to keep the loud passages, whatever their original Wide Track and Push-pull. By doubling intensity, whereas the light valve pro- the width of the track the noise would be duced a spot on the film of constant magnitude, at nearly full track amplitude. (See section on “Stereophonic.”) expected to increase 3 db but the use- intensity but varying height, and there- fully modulated light would increase 6 for varied the time of exposure. Since initial consonants usually carry little power compared with vowels, the db. While the improvement is not The light valve also was modified to compressor may wait to act for the large, it is worth while, and has been make variable-area records by turning it vowel, with resulting relative exaggera- widely employed for original recordings 90’ so that the ribbons were vertical tion of consonants (especially sibilants). and for special purposes, when the sound (parallel with the direction of film This has been corrected by a high- print does not have to carry a picture. motion) and the motions of their edges frequency pre-emphasis in the modula- Push-pull systems have also been imaged at 10 : 1 magnification on the tion which controls the compressor.301--303 developed for both and area183 film. This was the system employed for Anticipation in compressor systems is systems and for wide tracks310z311and the stereophonic dernonstration~.~08~295~296 desirable (see section on “Ground- standard-width tracks. Standard-width Variable-area optical systems using light Noise Reduction”). Fast action of com- push-pull systems have been considered valves were shortly thereafter offered as pressors is essential and avoidance of for general theater reproduction, but optional equipment to licensees. “thump” is one of the problems.305J06 their actual use has been limited to By using valves having one, two or Experiences in recording with com- places where the required special re- three ribbons as needed, the light-valve pressors are reported by Aalberg and producing systems did not involve large system has been made to produce many Stewart of RK0300. Experience with investment. The push-pull system does of the various types of area track that not give any improvement in signal-to- are produced by the RCA galvanometer * Probably later adopted magnetic recording. noise ratio except for the few film

Kellogg: History of Sound Motion Pictures 427 blemishes or other disturbances which operate on the recording amplifier to was described by Frayne and Her~-nfeld.~l~ affect the light through both sides alike. change its gain as in any automatic Between the normal soundtrack and The principal advantage of the push- compressor, and the same voltage can picture areas, space was found for a pull system is that it reduces distortion, operate simultaneously on the control soundtrack 0.005 in. wide. The authors and in the density system it may thereby tone to make the appropriate change in give their reasons for believing that somewhat increase the permissible ratio either its amplitude or frequency. In frequency modulation of the control tone of light modulation. As applied to area reproduction the output of the control- would be more reliable than amplitude recording it almost eliminates zero-shift tone track is then used to provide a modulation. With a frequency range of distortion. With both systems, push- voltage which is directly related to that one octave and using a bandpass filter, pull operation permits the use of faster- which altered the recording gain, and they found that in spite of the narrow- acting ground-noise reduction, since can therefore be used to produce inverse, ness of the track it provided a 38-db the “thump” caused by rapid change in or compensating, changes in the repro- signal-to-noise ratio. The system was transmitted light is largely cancelled. ducing gain. designed to afford changes of gain for the Milestones in the progress toward better The use of an extra soundtrack or sound up to 30 db, thus expanding the sound are usually fixed in our minds by recording to be used in the manner just dynamic range by that amount. The major demonstrations. Such a demonstra- described was, I believe, first proposed soundtrack was of the standard-density tion was given at the 1935 spring con- by C. F. Sacia of the Bell Laboratories, type. vention. Using wide-track push-pull and described in U.S. Pat. No. 1,623,756. In certain systems in which three or density recordings, improved two-way One of the first in the RCA group to more independently recorded sound- (divided-frequency-range) loudspeakers become interested in control tones was tracks are used, the sound and picture and amplifiers with abundant reserve Charles M. Burrill who experimented are on separate films. This gives plenty power, the engineers of the M-G-M with tones superimposed on the modula- of room for the control track, but the Sound Department, under the direction tion, but of either too high or too low practice has been to allot only one track of Douglas Shearer, gave impressive frequency to be reproduced by the audio to the control and to superimpose the demonstrations. system, especially a subaudible tone tones, separating them in reproduction Push-pull wide-track is regarded as such as 20 cycles. For film recording he by bandpass filters. the last word in photographic recording proposed scanning the sprocket-hole and has for years been used in many area, and varying the magnitude of the Class B Push-pull. Of the possible studios for original recording. resulting 96-cycle tone by blackening systems of photographic recording, the If the original recording is wide- more or less of the film between per- Class B area track undoubtedly carries track push-pull, the positive (which is forations. Thus, if these areas are left furthest the principle of low print trans- edited and then used as the master for clear, the 96-cycle tone is comparatively mission when the modulation is low. re-recording to the release-print nega- feeble, while the maximum is produced Such a system was described by Dimmick tive) would be a print of the original. if they are black. and Belar in 1934,314 and favorable If the original is magnetic, the favored The sprocket-hole control track was experience in its use was reported in practice has been to re-record to a wide- developed for application by H. I. 1939 by Bloomberg and Lootens of track push-pull direct positive for the Reiskind and adopted by Warner Republic Pictures,315 where it had been edited film. In eliminating the step of Brothers for their “Vitasound” system, adopted for original recordings and been printing, the direct positive minimizes which will be described in the section on used in the making of fourteen pictures quality losses. multiple-speaker systems.326 at the time of the report. They have Control tones of the subaudible type continued to the present time using the Control T0ne.3’~ The most effective have in recent years found use for pro- Class B push-pull system for all original way to reproduce the great range of ducing spread-sound and stereophonic recordings. They also reported the sound volume encountered in natural effects which are particularly appropriate methods used for test and adjustment. sounds is to resort to compression for the for accompanying large-screen pictures. In the Class B system one side of the recording and to expand in reproduction. (See Perspecta Sound, under “Multiple- track carries only the positive parts of It has always been the practice for the Speaker Systems” below.) The chief the waves and the other only the nega- recordist to use his controls to maintain advantage of this method of recording the tive. When there is no modulation there the recorded levels between the limits control tones is that it requires no is no clear film, and no ground-noise- set by overload and a satisfactory margin changes in the recorders and scanning reduction system is needed ; therefore above noise. Were a record made of each systems, the changes being confined to the transient or initial clipping which change in recording gain, and the pro- the electrical circuits. It meets the ground-noise-reduction systems can pro- jectionist given a cue sheet by which he important practical requirement that duce are avoided. Two triangles of light could at the right times make the inverse reproduction must be acceptable on (see Fig. 13) are formed at the plane of changes in reproducing gain, the natural standard equipment (with no provision the slit (one for each half of the track) levels could be restored. Such a method for control-tone use). Variable-area and their vertices just touch the slit of operation was at one time contem- recording is better suited than variable- with no current in the galvanometer. plated. However, manual controls in re- density for use with subaudible control Exposure takes place only as one or the cording are too slow, and manual restora- tones, since with suitable processing other triangle illuminates more or less tion of level too unreliable. Automatic more accurate linear relationship be- of the slit. A feature not described in control of gain on the other hand has tween input and output can be main- the papers just mentioned is shown in a been used very effectively. If space can tained throughout the range of modula- September 1937 paper by Dirnmick.’= be found on the film for an extra sound- tion, thus avoiding modulation of the In the mask which is imaged at the plane track, a continuous tone can be recorded audio by the control tones. The maxi- of the slit to form the triangular light with either its amplitude or its frequency mum level which can be recorded must spots, a narrow slit extends from the automatically correlated with the gain of be reduced 2 or 3 db, to make room for vertex of each triangular opening, so the recording amplifiers, so that it the control tones, so that the sum of the that light is never completely cut off provides a complete record of the two will not exceed the permissible at this point. There is thus formed a recording gain throughout the recording. maximum. Tones of 30, 35 and 40 cycles continuous line at the middle of each For example, a voltage derived from have been used for the controls. half track, not wide enough to let an the modulation to be recorded (and A control-tone system used by ad- appreciable amount of light through thus a function of its initial level) can herents of the variable-density system the film, for it is so narrow that it is

428 August 1955 Journal of the SMPTE Volume 64 largely fogged in, but it prevents a slight line, covering a large range of trans- is a definite practical limit set by general wave-shape distortion which might other- mission. A method of correcting in theater noise.* wise be produced by image spread, just the reproduction for the nonlinear trans- Multiple-Speaker Systems as the tip of the triangle crosses the slit. mission characteristic of a variable- density negative was described by Alber- It has often been observed that musical Direct Positives. Recording direct posi- sheim in 1937,318 but this would limit reproduction gives greater satisfaction tives, instead of recording negatives and the usefulness of the direct positives. if it comes from several sources. The then making prints from which to make Electrical compensation in the recording Music Hall at Radio City in New York re-recordings, would seem to offer was described by 0. L. Dupy of M-G-M has been equipped to make this possible, important advantages in simplifying in 1952.319 An approximation to the and thus has afforded numerous oppor- operations and reducing time loss. desired relation is possible by recording tunities to verify the advantages of There would also be an advantage in on the toe of the H&D curve, but high- multiple sources. With reference to the reduced ground noise since one less level output is not possible from such effort to obtain “dynamic range,” film is used and thus one less source of direct positives without serious dis- Garity and Hawkinsa3 state that: “Thrce graininess. The high-frequency losses tortion. channels sound louder than one channel due to the printer would also be avoided. A radically different solution of the of three times the power-handling In the section on “Ground-Noise Reduc- problem was described by Keith and capacity. In addition, three channels tion,” it was stated that a ground-noise Pagliarulo in 1 949.320Superimposed on allow more loudness to be used before reduction system applicable to variable- the audio current supplied to the valve the sound becomes offensive, because area direct positives had been developed was a 24-kc bias current of twice the the multiple source, and multiple stand- with the feature of anticipation, so that magnitude required to modulate fully the ing-wave pattern, prevents sharp peaks initial clipping need never occur. It light-valve opening at normal unmodu- of loudness of long duration.” was further possible under these condi- lated spacing. The ribbons are in That dialog should be reproduced on tions to work with closer margins, and different planes so that they can overlap a speaker as near as possible to where the so reduce the width of clear film. This without clashing. The authors reported actors are seen is never questioned, but system and results of tests with it are 8-db higher output from the direct- music and many sound effects such as described by Dimmick and Blaney in positive than from a standard-density thunder, battle noise and the clamor 1939,1g9with a further report by Blaney print, and 6-db higher signal-to-noise of crowds are far more impressive and in 1 944.316 ratio than a standard print without natural if coming from sources all The chief obstacle to making variable- ground-noise reduction. Direct positives, around the listener. The effects obtain- area direct positives had been that an generally push-pull, have come into able are discussed by H. I. Reiskinda6 entirely satisfactory way of avoiding wide use for editing and re-recording who also describes the equipment used in distortion due to image spread had not service.321,322 the sprocket-hole control-track system. been found. In the negative-positive Vitasound. This system, used by Warner process image spread in the negative Brothers, is described by Levinson could be compensated by that in the Electrical Printing. Successful recording and G0ldsmith.3~7It is the simplest of print, but the direct-positive system left of direct positives opens the way to the systems employing spread-sound no place for that solution. There is, to “electrical printing,” or putting sound- sources and control tone. It uses three be sure, a certain density for any film tracks on release prints by a re-recording similar loudspeakers, the usual screen- at which there is no image spread, and operation instead of by contact printing. centered dialog speaker and two side below which the blackened area is slightly This is discussed by Fra~ne.~~While speakers outside the screen area. The less than the exposed area. With the such a method would necessarily be more design of the system is based on the recording films available for a number expensive, the elimination of the flutter theory that the volume range which the of years, the density which gave no and irregularities resulting from the film (with the usual ground-noise re- spread, or zero shift, was so low that action of most contact printers is a strong duction) affords is adequate for dialog light passing through the darkened argument in its favor. Frayne finds and such other sounds as come from the areas would result in considerable noise. possibilities of greatly improved sound center speaker only, and that higher In some of the later fine-grain recording by this method, particularly for l6mm sound levels will be wanted only for films, however, the balance occurred at color prints of the reversible type such as music and sound effects for which the densities which were satisfactorily high, Kodachrome or Ansco-color. spread-sound source will also be wanted. thus making direct positives feasible Engineers have long been attracted The control-tone output is therefore under conditions in which they had not by the possibility of improved resolution used, first to switch in the side speakers, been before. However, the direct-positive and reduced distortion in 16mm positives with no change in total sound power. system described by Dimmick and by direct recording,324but until the more Further increase in control-tone output Blaney was not limited to the use of recent developments, direct positives raises the gain on all three speakers, up special films. They used a push-pull (except on reversal films) were not a to a total of 10-db increase. system which goes so far toward neu- success. tralizing the effects of image spread that Stereophonic Sound. Mention has been excellent results were obtained with the made in the section on loudspeakers of 4 standard fine-grain recording stock for Limits of Volume Range. While ideal the transmission of music and other variable-area.316J17 sound reproduction would seem to call sounds “in auditory perspective” from In variable-density recording the same for duplication of original sound levels Philadelphia to Washington in 1933. 4 anticipation feature is not applicable, it is questionable whether illusion or Three microphones were spaced across but the other advantages of direct posi- seeming naturalness is improved by the stage and their outputs separately tives for original recordings apply. The going as far as this in the direction of transmitted to three similarly located problem of recording a positive lies in loudness. In the applications of control- loudspeakers at the auditorium where the requirement that the film transmis- tone systems, where very high levels are the sound was reproduced. The various sion shall not be a reciprocal of the attainable, the extremes have been exposure (hyperbola) which a non- scaled down by 10 db or more. With * Background noise makes the useful sound seem reversing film tends to produce, but respect to the desirability of carrying fainter than sound at the same level in a quiet place. This holds even though the background must have an inverse relationship ex- reproduction to extreme low levels, noise is not loud enough to interfere nor even to pressible by a downward-sloping straight W. A. Muelleras has shown that there make the listener conscious of its presence.

Kellogg: History of Sound Motion Pictures 429 orchestral soundsseemed to come from the emphasis and post-equalization’’ of high- taken from great classics. In the initial appropriate places, and a moving source frequency components of the sound was orchestral recordings many microphones such as a man walking across the stage employed. The usual ground-noise-re- and separate recording channels were and talking seemed at the receiving end duction system by light-valve bias was used. Recordings were selected or mixed to move about. In 1941 and earlier,32sJm not used, being made less necessary by in the re-recording to obtain desired similar demonstrations were given, with the compressor-expandor (or “Com- effects such as predominance in turn of the difference that this time the sound pandor”) system. various orchestral groups (strings, brass was recorded and reproduced. As in the The October 1941 issue of the Journal etc.). Sound and picture were on previous case, every effort was made by carries discussions by Fletcher,329Stein- separate films. The final sound film the engineers of the Bell Telephone berg,330Snow and S~ffeI,~~~Wente and carried three 200-mil push-pull variable- Laboratories and Electrical Research MUller,’O4 and by Wente, Biddulph, area soundtracks, and three super- Products Inc. to minimize all forms of Elmer and Anderson,332 of various imposed variable-amplitude control tones distortion and to reproduce the sound in aspects and elements of the stereophonic on a fourth track. The amplitude of the the full dynamic range of the original. system. several control tones was determined by Since it appeared that for the immedi- Despite the unquestionable success of manual adjustments. ate purposes of the project the films could the stereophonic system in reproducing The theater equipment consisted of be maintained in good condition with the subjective effects of sounds coming three loudspeakers at the center and to respect to abrasions and dirt, and that from the sources located as seen on the the sides of the screen and additional under such conditions the advantage screen, and the impressiveness of the speakers at the sides and back of the in terms of signal-to-noise ratio is with musical and sound effects which it was auditorium. The Iatter could be brought the variable-area system, the recordings capable of handling, the motion-picture into operation by relays responsive to were of the variable-area type, made with industry made no immediate move to notches in the edge of the film. They an adaptation of light-valve optics. adopt or apply it. A factor which un- were used effectively for various sound The recorder and reproducing machine doubtedly militated against interest in effects and for the music of a large carried the film through the translation utilizing the stereophonic system was chorus. Abundant sound power and points on smooth (toothless) drums, on that only for the patrons near the front volume range were employed, the whose shafts were damped flywheels of of a theater did the screen subtend a volume range being readily obtained by the liquid-filled type described by Wente large enough angle to make the difference use of the control tracks. and Muller.81J04 between stereophonic and single-channel The final Fantasound equipment was On the basis that the film track was reproduction impressive. It appears to designed by RCA engineers. A special capable of giving a signal 50 db above be another case of a development ahead optical printer was used to print sound noise, while the orchestral music to be of its time. With the recent advent from separate negatives onto the multi- reproduced had a range of 80 db, com- (commercially) of the wide-film systems track positive, and special film phono- pressors were used in recording, and CinemaScope and Cinerama, stereo- graphs in which a single system illumi- expandors in the reproducing system phonic sound, supplemented by “spread nated all four tracks, the transmitted light designed to compensate exactly* for sound,” plays an essential role in provid- being received by four independent the compression in recording. The ing the desired overall effects. double-cathode photocells. compressors made no change of gain until the signal neared full track level Fantasound: Fantasia. Walt Disney Fantasia was enthusiastically received and above this made the gain the inverse and his engineers had a somewhat but was not a financial success because of the sound level, thus keeping the different idea of what might be accom- of the heavy expense not only of its recorded level just below track over- plished by means of multiple-track record- production but of “road-showing’’ with load. The level of the amplified micro- ings with control tones. It might be such elaborate equipment. The advent phone output controlled the magnitude appropriate to say that they proposed of World War I1 hastened its with- of an oscillator tone, which tone was to make their spread-sound effects an drawal, but Disney had performed a simultaneously recorded in a fourth art rather than a science. Duplication great service to the industry and art by track and applied to the recording- of an original distribution of sound pioneering in a sound-effects field which circuit compressor. In reproduction the sources was a secondary consideration, is now finding important applications. same tone controlled the expandor gain. and the choice of directions from which Because the levels at the three micro- sounds were to come was to be entirely at Recent Multiple-Channel Systern~.%~-~~~ phones did not necessarily rise and fall the discretion of the directors, musicians During 1952 and 1953 three new systems together, the compressors in the three and technicians. of presenting pictures were introduced, channels were independently controlled The story of the development of known as “3-D” (Three-Dimensional), by their own modulation levels. This Fantasound and its evolution through “Cinerama” and “CinemaScope.” All necessitated use of three control tones, numerous experimental forms beginning of these represent more elaborate and but these were recorded superimposed in in 1937, was told by Garity and Haw- expensive picture-projection systems, and one track arid separated in reproduction kins,= with further reports by Garity with them would logically go whatever by filters. and Jones,%* and by E. H. could be offered in the way of more ! In order to give the compressor time to Garity and Hawkins reported from tests impressive sound. The principles of operate when a sudden increase of sound that if a sound source (from loud- stereophonic sound, plus the surround level occurred, an “anticipation” system speakers) was to seem to move smoothly speakers distributed around the audi- was employed, using two microphones. from one position to another, the output torium (as used in Fantasound), have The compressor operation was deter- power from the two speakers should be been applied. mined by the output of the microphone held constant. This condition is not In the 3-D system two projectors are closer to the source, while the second necessarily met by having the actual required, and a third synchronized microphone furnished the sound to be source move from near one microphone machine, a film phonograph, is added for recorded. to near the other, but it can be met when the sound. The sound system is stereo- As further insurance against audible the gain is reduced to one speaker and phonic and uses three 200-mil magnetic background noise, the system of “pre- increased to the other by means of a tra~ks.”OJ~~ knob or control tone. The CinemaScope systema9 (developed * Exactly, in terms of timing, but not necessarily The animated picture was designed by Twentieth Century-Fox) presents a fully restoring the 30 db of compression range. specifically for the music, which was much wider screen picture than does the

430 August 1955 Journal of the SMPTE Volume 64 standard system, and therefore presents in a number of theaters in Texas and sented great difficulties. Only the fact an excellent opportunity for the use of Oklahoma. It is described in the May that standards were much less exacting stereophonic sound since the screen 1954 Progress Reporta9 as using a stand- made the project practicable. But with speakers are distributed over a large ard optical track, but with provision each advance in the resolution of high enough distance to make the shifts in for shifting the sound from the center frequencies in 35mm recording, the the position of the sound source notice- speaker to either of two screen speakers corresponding principles were applied to able and needed for realism. The sound to right or left, or to peripheral (sur- 16mm. One factor at least was in the system developed to go with Cinema- round) speakers. This is accomplished favor of the low film speed, namely the Scope uses four magnetic stripes or tracks, by “switching cues in the form of a binary providing of adequate exposure. By the two just inside the sprocket holes and code marked into both sprocket-hole time of the outbreak of World War I1 two outside. To make more room for the areas,” and optically scanned. “Standard sound on l6mm film had improved to tracks and leave as much as possible for single-track optical release prints are the point which made it acceptable for a the picture, sprocket holes of new shape cued for use with this system by the great number of military applications.351 have been designed. In the new per- Dorsett representatives.” (Quotations Recording sound on black-and-white foration no valuable feature is believed are from the Progress Report.) reversal film was not too satisfactory, to have been sacrificed, and the film will but years later when color films came still run on standard sprockets. Three into wide use, it again became necessary and Color. Since in color pictures of the tracks provide the regular stereo- Sound put the sound on what was essentially the silver is removed from the emulsion, to phonic sound (three speakers behind a reversal type of film. Particularly special handling of the soundtrack is the screen) and the fourth carries sound successful for this purpose is recording on required. In the Technicolor imbibition effects for transmission by the surround the individual prints or electrical print- process the soundtrack is printed in the speakers. For multitrack magnetic ing. (See the section with that heading, usual way, the remainder of the film is reproducing equipment see references and the December 1950 paper by then cleared of all silver, and the pictures 108-1 10. J. G. Fra~ne.~~~) Cinerama3” (developed by Fred Waller produced by dye transfer from relief masters. Black-and-white prints, with original and Hazard E. Reeves, and commer- sound recorded either on 16 or 35mm cialized by Cinerama Productions Gorp.) With the multilayer color films such as Kodachrome, Kodacolor, Anscocolor films, formed the great bulk of the product carries the wide-screen principle much before and during the war. Fortunately, further, employing three projectors to and a number of more recent types it would be possible to expose the track fine-grained emulsions could be used. project adjacent edge-blended pictures Cost considerations dictated for the on a curved screen. For the sound six with white light so that all three dyes contributed to the density, or with most part the use of 16mm negatives magnetic tracks are used, and five and contact prints, and unhappily speakers behind the screen, while the colored light so that the sound image would be principally in the top layer, nearly all of the printers were of the sixth track feeds a set of surround sprocket type. One laboratory adopted speakers. and reproduce with a filter of the complementary color. The principal problem 1 : 1 optical printing as its standard.* The surround speakers, wherever these Much study was given to the possibilities are used, are of the direct-radiator type, with this procedure is that existing projectors are nearly all equipped with of recording direct positives. Still more much less bulky than the screen speakers development work was devoted to direct and with more limited frequency range. caesium photocells whose sensitivity extends well into the infrared, in which optical reduction from 35mm negatives, and this certainly helped resolution and Recent Multiple-Speaker Systems With Pho- range all of the dyes in use are transparent.3” Substitution of phototubes of high-frequency but in the tographic Track.a9 A simpler multispeaker case of area tracks, neutralization of (or sound-placement) system has been other types would mean an objectionable loss of sensitivity. In May 1946, A. M. image spread in the positive by spread developed by Robert and named in the negative was not easily achieved.n6 “Perspecta Sound.” Mention was made Glover of RCA reported the development of a blue-sensitive phototube which The same factor made difficulty in the under the heading “Control Tones,” recording of direct positives, which had of the use of control tones of subaudible could replace the caesium tubes in projectors without loss of output.3“,“5 had advocates since early in the era of frequency. In the Fine system there is sound pictures. Since 35mm negatives only one soundtrack (variable-area, However, by this time the makers of color film had begun the use of edge- for optical reduction, or 35mm re- photographic) in the standard position, recording positives to be used for making but three loudspeakers behind the development processes that form a track of silver or of a metallic salt (opaque to l6mm negatives give best results if screen, fed through separate variable- recorded with different degrees and gain amplifiers. Three control tones, infrared) not removed by the bleach which was part of the picture process- kinds of high-frequency pre-equalization, superimposed on the sound recording the organizations producing these masters and separated in reproduction by filters, ing.346In one system the silver is in the top layer only, and the dyes in the other’ have worked together to establish stand- control the gains of the three amplifiers ards of recommended characteristic^.^^^ SO that the apparent source can be made two layers contribute to the opacity. to shift across the screen, and the total The metallic track became general in Adding sound to 16mm films by use sound level may be varied as desired the handling of sound on color film~.~~’-35’J of a stripe of magnetic material promises to become of great popular and com- 4 to increase the dynamic range. The mercial importance.3” Perspecta-Sound recording if repro- Sound on 76mm Film. I have confined duced on a standard system (not my story to developments in the 35mm Drive-in theater^.^" The drive-in equipped to make any use of the control field in the belief that the general prin- d theater was first proposed and advocated tones) would, except for a slight reduc- ciples and practices which solved prob- by R. M. Hollingshead, a businessman tion in level, be indistinguishable from a lems or led to improvements in one case of Camden, N.J., not affiliated with standard recording. M-G-M and Para- were applicable in the other. It is prob- motion-picture or electronic interests. mount have been particularly interested ably a gross injustice to pass over the The first such theater was built near in Perspecta Sound and have equipped valuable work of engineers whose efforts Camden in 1933. In the earlier experi- numerous theaters with it.a2 were devoted to l6mm sound, but it ments with the system effort was made Another sound-placement system was seems to be a necessity. developed by the Dorsett Laboratories of From the start, the recording of * Precision Film Laboratories, New York, oper- Norman, Okla., and has been installed sound at 36 instead of 90 ft/min. pre- ated by J. A. Ma~rer.2~0

Kellogg: History of Sound Motion Pictures 431 to put out enough sound power from a Erasure was by a saturating d-c field, progress the result of laborious experi- screen speaker to enable patrons to hear and a d-c bias was combined with the mentation and wise application of satisfactorily in their cars. This presented recording-voice current in sufficient refinements and better techniques. Not great technical difficulties, and also magnitude to maintain all the recorded until a sound quality was attained in would have restricted theaters to loca- magnetization of the reversed polarity. magnetic recording surpassing that of any tions where the noise would not be too An alternating magnetic field has for other known system did it become of objectionable. Several arrangements were years been the accepted means of great concern to motion pictures. tried, one with loudspeakers distributed demagnetizing magnetic materials, and Development of magnetic coatings on over the field so that each speaker it seemed a paradox that a superimposed paper or other base materials was under- provided sound for two cars side by side. alternating current would assist in, taken by the AEG in Germany about This was a great improvement from the rather than obliterate, recording. How- 1928, but up to the outbreak of war noise standpoint, and the theaters pre- ever, it was found in 1921 by W. L. nothing of outstanding quality had viously equipped with screen speakers Carlson and G. W. Carpenter, then appeared. At the close of the war the were converted. However, these “out- working on a government project of American occupying forces brought car” speaker arrangements still left recording telegraph signals, that a bias back samples of a new German magnetic something to be desired on the score of current of supersonic frequency could tape and equipment. The magnetic general noise. The “in-car’’ speaker, be very advantageous in magnetic material was a finely divided iron oxide, introduced by RCA in 1941, provided recording of audio-frequency signals.358 mixed with a binder and coated on a sound which was much more satisfactory The development of magnetic record- thin cellulose base (total thickness to patrons and practically eliminated ing was carried on by the Bell Telephone about 0.002 in.). In the recording magnet the neighborhood-noise problem. In the Laboratories during the 1930’s. A thin a supersonic bias current was super- design of “in-car” speakers, the qualities steel ribbon about + in. wide was imposed on the audio current. Magne- of ruggedness, conveniently small size found better than wire, and alloys of tization was longitudinal, produced by without too much sacrifice of sound superior magnetic properties were de- C-shaped magnets of high-permeability quality, and immunity to damage by veloped. Direct current erase and bias alloy, with very short gaps where they weather were design 0bjectives.~55The were used in the equipment. contact the tape. Reproducing magnets amplifier and audio-power distribution The developments and publications of were similar. In cleanness of reproduc- system had to be such that the individual Marvin Camras359*360of Armour Re- tion, low ground noise and volume range levels would not be too much affected search Foundation aroused widespread the German system set a new high by the number of speakers in use. interest in magnetic recording and the standard. During the 1930’s a number of drive-in results obtainable with high-frequency A period of intensive development theaters were built in New England, bias. The essential difference between a followed the demonstrations. Some time in the South, and (through the efforts recording made with d-c bias, and one was required before American firms of Philip Smith) in Indianapolis, Cleve- made with high-frequency bias is that could match thc quality of the German land, Detroit, Milwaukee, St. Louis, with the latter (1) modulation is through tape, especially with respect to freedom Kansas City and Pittsburgh. Building zero, with resultant increase in maximum from noise. Numerous papers were pub- activities were practically at a standstill recordable amplitude, and (2) there is no lished describing basic properties of during the war, but after its close the remanent magnetism when the modula- magnetic materials,361-362 analysing the number increased rapidly. Underhil1356 tion falls to zero. The second feature is action of the supersonic bias,363J64and gives the number of drive-in theaters at the condition for low background noise. reporting tests on various tapes to the end of the war as about 60 and four The comparison is analogous to that determine optimum bias, relation of years later as 1000. He attributed the between a Class B push-pull photo- distortion to recorded amplitude, and rapid growth just after the war in part graphic record, and a unilateral record- amount of residual noise.365Among the to the greatly increased number of cars ing without ground-noise reduction. first in this country to produce acceptable on the road, especially after the end of There was much developmental activ- tapes was the Minnesota Mining & gasoline rationing, and in part to the ity in wire recording just before and Mfg. CO.~~~Experimenters tried the shortage of building materials which during the war, with numerous applica- magnetic black oxide (Fe304) in powder made the construction of indoor theaters tions for the armed services. The extreme form and found it capable of producing relatively difficult and expensive. The compactness of recordings stored as higher output levels but noisier than the May 1954 Progress Report puts the magnetized wire plus the ruggedness of red oxide used in the German tapes. number of drive-in theaters at the end of the system with respect to mechanical The red oxide (Fe20a) in a certain 1953 at 4000, and estimated that six injury made it of especial value for mili- crystal form (“gamma phase”) is mag- months later there would be 4500. tary uses. Better wires were developed, netic. Grinding to a particle size of The most serious problem of the drive- 4- to 5-mil diameter sizes being widely about 1 p or less was found essential, in theater is the provision of sufficient used. The Brush Development Co. de- and prolonged mixing. Many tape- light for the large screen, a consideration, veloped a compound wire in which an recorder-reproducer equipments were however, which lies outside the scope of alloy of superior permanent magnetic developed, both for amateur and pro- this paper. properties was plated on a brass c0re.3~~ fessional use, the latter principally for The National Standard Co. of Niles, broadcast stations. Magnetic Recording Mich., developed a successful stainless- The principle of magnetic recording steel recording wire. Advantages of the Magnetic System. The was demonstrated in 1900 by Valdemar S. J. Begunls7 quotes an article in compactness and portability of the mag- Poulsen, who called his device the Machinery of January 1917 about the netic-tape equipment, its freedom from “Telegraph0ne.”27*35~About 1917 the Telegraphone in which mention is made dependence on laboratories, the immedi- American Telegraphone Co. marketed a of a stripe of powdered iron to be painted ate playback, the small storage space dictation machine under that name on a motion-picture film to provide required, the relatively small cost of which performed well by existing stand- synchronous sound. It was more than recording stock, plus the ability to re- ards but was complicated and could 30 years later before use was made-of use tape when the recorded sound on it not compete commercially with cylinder this principle - one of the many ex- is no longer wanted, the ability to work machines. Recording was on 0.010-in. amples of development in which the basic in daylight, and finally the excellent “piano” wire, contacted from opposite concept is only a small part of &e sound quality and dynamic range, sides by a pair of offset steel pole-pieces. invention, and the real contribution TO combined to make the magnetic system a

432 August 1955 Journal of the SMPTE Volume 64 great advance from both the economic and Johnson365of $4.. thin tapes made photographic-sound records for this and performance standpoints. by Du Pont and by Minnesota Mining & intermediate function. Push-pull re- Mfg. Co. as compared with one of the cording is usual, and in order to add as Uses in the Motion-Picture Industry of German tapes. The commercial version little ground noise as possible, prefer- Nonsynchronous Recording. The prompt of the conversion kit wassdescribed ,by ably wide-tmck. The use of direct posi- ’ interest of motion-picture producers in Gunby3e9at the October 1948 convention. tives saves time as well as printing losses. magnetic recording was shown in two In the June 1949 Journal Mueller and When the editing is completed, the lines of activity. The Basic Sound Com- Groveszo0describe experiences at Warner negative for release printing is made by mittee of the Motion Picture Research Bros. Studios with various uses of non- another re-recording operation. Council held meetings in 1946 to which synchronous magnetic recording, and To provide the visual advantages of prospective suppliers of magnetic-re- such experience as had been had up to the photographic track for purposes of cording equipment and record materials that time with synchronous magnetic editing while retaining the quality were invited.366 The purpose was to recording. The practice of re-recording advantages of the magnetic system, an formulate the requirements of the from magnetic to direct-positive photo- arrangement has been tried which motion-picture industry in order that graphic tracks is mentioned. This has registered by an ink line on the magnetic developments might be better directed sometimes been called “electrical print- film the amplitude of the modulation toward meeting these requirements. ing,” and soon became the prevalent being rec0rded.3~~Another expedient, The other activity was making experi- method of providing a sound film to be described by Frayne and Li~adary,~~~is mental use of the nonsynchronous thin- edited and then re-recorded to the final to make simultaneous tracks on the film tape recording equipment which was the release negative. The January 1952 to be edited, one magnetic and the other first to be developed.200While the most paper by Carey and M~ran~~~states photographic variable-area, the mag- important use which might be made of that the practice of re-recording from a netic to be used for the re-recording. the magnetic system would be for the magnetic original to a 200-mil push-pull A system in which both pictures and original synchronous recordings, there variable-area direct positive for editing sound can be backed up at any time are numerous operations for which the had been followed by Universal Inter- without loss of synchronism is described lack of exact synchronism of magnetic national Pictures since January 1951. by George Le~in3~~of the Signal Corps tape would not be too serious an obstacle, The Progress Report of May 1951 Pictorial Center. This makes it possible such as for immediate playback after shows RCA synchronous magnetic re- for a narrator to correct or change his rehearsals, for training singers and cording equipment, designed to make speech, erasing portions of the previous actors, and for recording of music and one, two, or three tracks on the film.* record as he substitutes the new. sound effects which do not have to This was being used by Columbia, the synchronize exactly with the picture. sound being re-recorded to direct posi- Synchronous Thin Tape. While use of In musical numbers the picture is often tives for editing. The same report perforated film afforded the necessary secondary to the sound, and in many tells of Westrex portable synchronous synchronism, the thin tape has the cases the practice had been followed of magnetic-recording equipment, suited important advantages of reduced size recording the sound first, without pic- (optionally) to 35mm, 17gmm or and weight of equipment, smaller space ture, and subsequently fitting the picture l6mm film, describing this equipment as required for storage, lower cost, and to the music. For this, initial recording on in wide use in this country and abroad. better contact between magnet and tape was applicable, the sound being Triple-track Westrex equipment is de- record surface because of its greater re-recorded to film while musicians or scribed by Davis, Frayne and Tem~lin~~lflexibility. A number of systems have, vocalists performed synchronously for in the February 1952 Journal. therefore, been developed to drive tape the camera. Portable magnetic-recording equip- in strict synchronism. These have for the ment (the complete channel weighing most part used the principle of recording Papers on Progress in Magnetic Sound. less than 100 lb) using 17grnm film a tone on the tape in addition to the audio At the October 1946 convention of this was described by at the April modulation and of controlling the speed Society Marvin Camras described and 1950 convention. The May 1951 Progress of the driving system so as to hold demonstrated synchronous magnetic Report states that since April 1, 1950, all within limits the phase relation between sound, using 35mm film coated at Paramount production recording had the reproduced tone and a reference Armour Research Foundation. However, been done on this equipment. frequency, usually the 60-cycle power the coated film was not offered com- which drives the camera. The recorded mercially. At the same meeting H. A. The Progress Report373of May 1952 tone may be 60 (excluded from states that by the end of 1951 approxi- Ho~e1136~of Indiana Steel Products the audio by a high pass filter) or a tone mately 75% of the original production Co. gave data on several magnetic slightly above the audio range for recording, music scoring and dubbing in materials and described a coated-paper example 14,37615377 or 18378 kc, and Hollywood was being done on magnetic- tape which could be perforated if modulated at the 60-cycle rate. recording equipment. desired. While paper offered certain An optical tone on the tape, as advantages which commended its use by stripes on the back of the tape at the time, it has not been possible with would have certain advantages (one of coatings on paper to attain the quality Editing. While it is entirely possible to which is that full voltage can be de- 9 and low noise that are realized with edit magnetic recordings with the help of veloped from standstill), but this has to bases of clear plastic. quick-stop reproducers, sound-film edi- the best of my information not been In February 1947 the Du Pont Co. tors have come to depend in part on the made commercial. furnished RCA with samples of coated i visibility of modulation in a photo- Experiments have indicated that with 35mm film base. A conversion kit was graphic track. This is probably one of a suitably designed mechanical system, developed so that a standard RCA the chief reasons for the retention of perforations and sprockets are not out photographic recorder (PR-23) could of the question, even with tape as small record and reproduce magnetically. *The use of three tracks on a 35mm film was as 0.002 in. thick by in. wide.379 The November 1948 issue of the Journal begun by Columbia in November 1950. In single- contains a description of the kit by channel recording the three tracks could be used Striping. While magnetic sound has Ma~terson,~~~measurements of the prop- in succession, or in recordings calling for combined voice, orchestra and sound-effects these been thus far used mainly in the pre- erties of the Du Pont film reported by could be independently and simultaneously re- liminary operations of making sound O’Dea,368and bias studies by Dimmick corded and later mixed for desired balance. pictures, and release prints are still

Kellogg: History of Sound Motion Pictures 433 optical, there are some exceptions, for the safety base were so inferior to the In the early days of the Fleming valve and example the films for the Cinemascope nitrate base that it was not a satisfactory the de Forest audion the opinion was widely system, on which the sound is all- substitute. held that these depended for their conducting properties on the prewnce of some gas. The film companies worked long and magnetic. This calls for applying stripes Experiments in 1907 by Prof. Soddy396 of magnetic coating on the photographic diligently on the problem of improving of the University of Glasgow lent support film which carries the picture. There are film-base stock. In 1937 the Eastman Co. to the belief that the space current would ~ also numerous applications, especially adopted an improved safety base,390and in become zero if a true vacuum were ob- with 16mm films, for which it is desir- 1948 announced a new safety base which tained. On the other hand, the English I able to add sound to an existing picture combined the needed properties to scientists J. J. Thompson (who first proved film, and other uses for which the stripe replace the long-used nitrate base, being that there was such a thing as an electron) is applied to the unexposed film. superior to the nitrate in heat resistance and 0. W. Richardson believed that high The application of the magnetic and low shrinkage. It is described by temperature would cause a conducting body to throw off electrons into the adjacent F~rdyce~~l.With these virtues the new material in stripes of closely controlled space, without dependence on the potent width, position and thickness is no film has rapidly supplanted nitrate. effects (in rcleasing electrons) of impacts simple problem. The April 1953 Journal, The improvement in sound quality on the cathode surface by positive gas ions. Part 11, carries a series of papers dealing due to reduced shrinkage may not be It was the question which of these theories with problems of magnetic recording. noticeable to the average listener, but it was right that enlisted the interest of Dr. Striping is discussed by representatives must inevitably mean better performance Irving Langmuir. By using the known of Reeves Soundcraft Cor~.,~~~Minne- especially in the action of contact techniques for driving occluded gases from sota Mining and Mfg. CO.,~~~Eastman printers. The importance of this new glass and metal surfaces, the Gaede diffu- Kodak CO.~~~and Bell and base can hardly be exaggerated. sion pump, and adding a liquid-air cooled trap in the exhaust line, Dr. Langmuir The Minnesota Mining and Mfg. Co. Acknowledgments exhausted his experimental tubes to about I developed a method of application the highest possible vacuum. He showed of a stripe by transfer from a temporary I wish to express my appreciation and that (with adequate temperature of the supporting tape. Only heat (no solvent indebtedness to the many who have tungsten cathode) a pure electron space or wet cement) is required for the supplied information, and reviewed current flowed, and that this current transfer. The Reeves engineers also portions or the entire text, submitting followed the theoretically predicted 1 .5 have papers on the preparation of the comments and indicating corrections. power relationship to the anode potential. 397 magnetic and the study of Special acknowledgment must be made With high enough anode voltage to carry the minute surface irregularities which to E. C. Wente who supplied me with over all the electrons emitted by the cathode tend to lift the tape from the magnet much of the story at Western Electric, (“saturation current”) Langmuir was able causing sound “drop-o~ts.”3~~In spite and to J. G. Frayne who reviewed a large to verify 0. W. Richardson’s prediction of the utmost effort to prevent the part of the text, indicating corrections (“The electrical conductivity imparted to a vacuum by hot conductors,” Phil. Trans. formation of such high spots they are not and answering many questions. A. C. 207: 497, 1903) that the rate at which entirely eliminated, and polishing opera- Blaney gave particularly generously of electrons are “boiled” out of the cathode tions are helpful. Other papers in the time and thought in answering questions bears a similar relation to temperature fc) group deal with wear on magnets,3s6 and filling in gaps of information. My that vapor pressure does in the case of an OIW measurements of magnetic induction,3s7 thanks also go to E. M. Honan, and evaporating substance. The presence of and standardizati~n.~~~ W. V. Wolfe, 0. B. Gunby, E. W. small quantities of ordinary gases (other The May 1954 Progress Report men- Templin, Kurt Singer, Barton Kreuzer, than the “noble” gases) was found to tions new high-output magnetic-oxide and Walter L. Tesch for reviews and “poison” the tungsten surface and greatly coatings introduced in 1953 and 1954 comments. reduce emission, but this effect disappeared at high enough temperatures.397 by Minnesota Mining and Mfg. Co., greatly increased use of striping, new Thesestudies were begun in August 1912, machines for applying the stripes, and APPENDIX and continued through that year and the following. Three electrode tubes exhausted several designs of theater soundheads Development of High Vacuum Ampli- to high vacuum were found to be free from (to be mounted above the projector fier Tubes from the Audion the voltage limitations and erratic behavior head) for reproducing multiple-track of the previous audions.398 Another impor- magnetic sound. A brief account of the improvement of the audion by Dr. Arnold of the Western tant outcome of the development of the A Sound Committee Report by J. K. Electric Co., and of the subsequent patent pure electron discharge was the Coolidge Hilliard in the June 1953 Journal tells of interference and litigation, appeared in X-ray tube (hot-cathode, high-vacuum arrival at agreement for a standard of Jour. SMPE, 17: 658-663, Oct. 1931. type) in which the electron velocities at the track positions for triple 200-mil magnetic In view of the vital part played by the anode (and thereby the frequency or tracks on 35mm film, of projects for electronic triode, it seems appropriate to penetrating power of the X-rays) could be accurately and reliably controlled, and standardizing theater magnetic-repro- tell something more about the parallel developments at the General Electric carried to much higher values than had ducing characteristics, and plans for been possible in the previous tubes where making available various magnetic test Co. The results, in terms of practical, high-vacuum amplifier tubes, were attained gas ionization had limited the effective films, corresponding to the long-used independently and at very nearly the anode voltage. The use of chemical photographic test films. “getters” for improving the vacuum in same time at the two laboratories. The I Theater reproducing systems for mag- Telephone Co. had immediate use for sealed-off tubes was also much advanced + netic tracks are designed to be mounted audio amplifiers and promptly put them by work at this time at General Electric. on the tops of projectors, and do not to use on a large scale, whereas, at the In October 1913, a patent application was interfere with the optical reproduction. General Electric Co., the work was much filed for Langmuir on “Electrical Discharge more nearly a pure research. Apparatus” (triode) in which conduction I~ Study at the Research Laboratory of the is entirely by electrons, the effects of gas New Safety-Film Base.2so~3sgFor years ions being negligible. the motion-picture industry struggled to General Electric Co. of the relation of to The audions which de Forest supplied for minimize the fire hazard of nitrate film. “Edison Effect” current residual gas was an outgrowth of incandescent lamp radio reception broke into a glow dis- Safety-film base had early been de- development. Because of serious effects in charge if anode potentials of more than veloped and used for certain purposes for some cases of the minutest traces of im- 20 to 40 were applied, and all control by 35mm film, and was mandatory for purities or of certain gases or vapors, the grid then vanished. The very limited amateur equipment (lbmm), but the the techniques of producing high vacua output with low-voltage operation was not stability and mechanical properties of had been developed to a high degree. serious for radio detectors, but made the

434 August 1955 Journal of the SMPTE Volume 64 tubes of little or no use as amplifiers. Some light on the questions at issue may 251. C. W. Faulkner and C. N. Batsel, “Opera- tion of the variable-intensity recording sys- During the years 1909-1912 de Forest was be found in opinions written by the succes- tem,” Jour. SMPE, 36: 125-136, Feb. 1941. employed by the Federal Telegraph CO. of sive courts. 252. W. C. Miller, “Speed up your lens systems,” California. The company wanted amplifier Court of Appeals of the District of Jour. SMPE, 35: 3-16, July 1940; and “Re- tubes and de Forest ordered some made Columbia (339 Oficial Gazette 56). cent improvements in nonreflective lens with better vacuum, calling in some cases Patent issued pursuant to finding of coating,” Jour. SMPE, 37: 265-273, Sept. 1941. for re-exhaust, so that by August 1912, a this court. tube was being used at 54 v and by Novem- 253. C. H. Cartwright and A. F. Turner, Federal District Court for Delaware ber, one at 673 v. This was held in some of “Multiple films of high reflecting power,” before which the General Electric the later court actions to be a clear indica- Phys. Rev., 55: 595, June 1939. vs. de Forest suit was tried (Federal tion of the direction from which improve- 254. G. L. Dimmick, “A new dichroic reflector Reporter, Val. 23, 2nd Series, p. 698). and its application to photocell monitoring ment in the audion could be expected and systems,” Jour. SMPE, 38: 36-44, Jan. 1942. thus, so far as invention or dircovery goes, U.S. Circuit Court of Appeals for the Third District (44 Federal Reborter 255. Mary Ellen Widdop, “Review of work on an anticipation of possible patent claims dichroic mirrors and their light-dividing by others, directed to improving the audion 2nd Serirs, p. 931, and U. S. Pat. characteristics,” Jour. SMPTE, 60: 357- by employing higher vacuum. Quarterly, 0ct.-Dec. 1930, p. 67). 366, Apr. 1953. Majority opinion for plaintiffs and 256. G. L. Dimmick and Mary Ellen Widdop, At the time that the work just described dissenting minority opinion. “Heat-transmitting mirror,” Jour. SMPTE, was going on, the Telephone Co. was U.S. Supreme Court (283 U.S. 664, 58: 36-42, Jan. 1952. making plans to establish transcontinental Baldwin Vol. XI 664). 257. Progress Reports, Jour. SMPTE, 56: 575 telephone communications in time for the May 1951; and Jour. SMPTE, 62: 336, opening of the San Francisco Panama- May 1954. Pacific Exposition in 1915. There had 258. G. J. Koch, “Interference mirrors for arc been a long-felt need for a voice-current References (Continued) projectors,” Jour. SMPTE, 55: 439-442, amplifier, much effort had been expended 232. J. Crabtree, “Sound film printing,” Jour. Oct. 1950. on the project, and devices based on SMPE, 21: 294-322, Oct. 1933, and 259. See reference 146 and accompanying various principles tried. “Sound film printing-11,” Jour. SMPE, classified bibliography. (J. I. Crabtree, 22: 98-114, Feb. 1934. (Repetition of ref. “The motion-picture laboratory,” Jour. For the 3000-mile transmission the 120.) SMPTE, 64: 13-34, Jan. 1955.) possession of amplifiers would be impera- 233. C. N. Batsel, “A non-slip sound printer,” 260. E. D. Cook, “The aperture alignment ef- tive and they must have low distortion to Jour. SMPE, 23: 100-107, Aug. 1934. fect,” Jour. SMPE, 21: 390-402, Nov. permit cascading. According to the ac- 234. A. V. Bedford, U.S. Pat. 1,754,187. 1933. counts by Lovette and Watkins,@ and by 235. F. W. Roberts and E. Taenzer, “Photo- 261. D. Foster, “Effect of orientation of the Wm. R. Ballard,393 a visit by de Forest in graphic duping of variable-area sound,” scanning image on the quality of sound re- Jour. 34: 26-37, 1940. produced from variable-width records,’’ October 1930 served to direct attention to SMPE, Jan. 236. R. V. Wood, “A shrinkage-compensating JOW.SMPE, 33: 502-516, NOV.1939. the possibilities of the audion, other devices sound printer,” Jour. SMPE, 18: 788-791, 262. N. R. Stryker, “Scanning losses in reproduc- of promise having till then claimed the June 1932. tion,’’ Jour. SMPE, 15: 610-623, Nov. research efforts. De Forest gave demonstra- 237. G. L. Dimmick, C. N. Batsel and L. T. 1930. tions and left a sample for tats and study. Sachtleben, “Optical reduction sound 263. E. D. Cook, “The aperture effect,” Jour. The demonstrations were repeated next printing,” Jour. SMPE, 23: 108-116, Aug. SMPE, 14: 650-662, June 1930. day for the benefit of research engineer 1934. 264. D. Foster, “The effect of exposure and de- 238. M. E. Collins, “Optical reduction sound velopment on the quality of variable-width OItow Dr. H. D. Arnold, who was quick to recognize the potentialities of the audion, printer,” Jour. SMPE, 27: 105-106, July photographic sound recording,” Jour. 1936. SMPE, 17: 749-765, NOV. 1931. the requirement for high vacuum and the 239. R. 0.Drew and L. T. Sachtleben, “Recent 265. C. N. Batsel and C. H. Cartwright, “Effect role of space charge in limiting and con- laboratory studies of optical reduction of uneven slit illumination upon distortion trolling electron current. He expressed entire printing,” Jour. SMPE, 41: 505-513, Dec. in several types of variable-width records,” confidence that an amplifying tube which 1943. Jour. SMPE, 29: 476-483, Nov. 1937. would meet the requirements could be 240. John A. Maurer, “Optical sound-track 266. American Standard, Scanning Beam Uni- developed from the audion, and he was printing,” Jour. SMPE, 50: 458-473, May formity Test Film for 35mm, Jour. SMPE, assigned the task. Progress was rapid, and 1948. 51: 542-544, Nov. 1948; and American tubes with much higher vacuum were 241. 0. Sandvik and J. G. Streiffert, “A con- Standard, Jour. SMPTE, 59: 427-431, tinuous optical reduction sound printer,” Nov. 1952. quickly available, but vacuum of the desired Jour. SMPE, 25: 117-126, Aug. 1935. 267. J. 0.Baker and R. 0. Drew, “New and old value was not achieved until a Gaede mo- 242. 0. Sandvik and G. Silberstein, “The de- aspects of the origins of 96-cycle distortion,” lecular pump, which was ordered from pendence of the resolving power of a Jour. SMPE, 37: 227-255, Sept. 1941. Germany, arrived. Long-lived cathodes photographic material on the wavelength of 268. J. G. Frayne and V. Pagliarulo, “Influence were of great importance for telephone light,” J. Opt. Sac. Am., 17: 107, Aug. 1928. of sprocket holes upon the development of applications, and cathodes of the Wehneldt 243. Carl L. Oswald and Warren D. Foster, adjacent sound track areas,” Jour. SMPE, oxide-coated type were developed to re- U.S. Pat. 1,928,392, 2,030,760, 2,055,261, 28: 235-245, Mar. 1937. place de Forest’s tungsten cathodes. 2,213,531. 269. J. Crabtree and W. Herriott, “Film per- 244. G. L. Dimmick, “Improved resolution in foration and 96-cycle frequency modulation More details about the developments sound recording and printing by the use of in sound-film records,” Jour. SMPE, 30: at the Western Electric Co. will be found ultraviolet light,” Jour. SMPE, 27: 168- 25-29, Jan. 1938. in the references already cited.4**29)Doubt- 178, Aug. 1936. 270. J. Crabtree, “Directional effects in con- ing the patentability of the improvement 245. G. L. Dimmick and L. T. Sachtleben, “Ul- tinuous film processing,” Jour. SMPE, 18: 207-231, Feb. 1932. brought about by higher vacuum, Arnold traviolet push-pull recording optical system for newsreel cameras,” Jour. SMPE, 31: 87- 271. J. Crabtree and J. H. Waddell, “Direc- and his attorneys did not file any applica- 92, July 1938. tional effects in sound-film processing-11,’’ tion until they learned that an application 246. J. 0. Baker, “Processing of ultraviolet re- JOW.SMPE, 21: 351-373, NOV.1933. based on a simiIar development had been cordings on panchromatic films,’’ Jour. 272. C. E. Ives and E. W. Jensen, “Effect of filed by the General Electric Co. The SMPE, 31: 28-35, July 1938. developer agitation on density uniformity prolonged interference which followed is 247. J. G. Frayne and V. Pagliarulo, “Effects of and rate of development,” Jour. SMPE, 40: summarized in the ultraviolet light on variable-density re- 107-136, Feb. 1943. (Repetition of ref. of the October 1931 cording and printing,” Jour. SMPE, 34: 158.) 273. M. S. Leshing, T. M. Ingman and K. Pier, conflicting opinions 614-631, June 1940. 248. C. R. Daily and I. M. Chambers, “Produc- “Reduction of development sprocket-hole U.S. Pat. No. 1,558,436 was issued to tion and release applications of fine-grain modulation” Jour. SMPE, 36: 475-487, Langmuir in 1925.” films for variable-density sound recording,’‘ May 1941. 274. 0. Sandvik and V. C. Hall, “Wave form In 1926 the General Jour. SMPE, 38: 45-55, Jan. 1942. 249. J. Strong, “On a method of decreasing re- analysis of variable-density sound record- suit for infringement against the de Forest flections from non-metallic substances,” ing,” Jour. SMPE, 19: 346-360, Oct. 1932. Radio Corp. Again there were decisions, J. Opt. SOC.Am., vol. 26, 73, Jan. 1936. 275. 0. Sandvik, V. C. Hall and J. G. Streiffert, appeals, and reversaL;~’&ding with the 250. W. P. Strickland, “An analysis oflow-reflec- “Wave-form analysis of variable-width ruling of the Suprem‘e Cbrik;May 25,1931, tion coatings as applied to glass,” Jour. sound records,” Jour. SMPE, 21: 323-336, that the patent did not involve invention. SMPE, 49: 27-36, July 1947. Oct. 1933.

Kelhgg: Hibtory of Sound Motion Pictures 435 276. C. E. K. Mees, “Some photographic as- area film recording channel,” Jour. 328. J. P. Maxfield, “Demonstration of stereo- pects of sound recording,” Jour. SMPE, 24: SMPE, 47: 299-326, Oct. 1946. phonic recording with motion pictures,” 285-326, Apr. 1935. See page 324. 303. Kurt Singer, “High-quality recording elec- Jour. SMPE, 30: 131-135, Feb. 1938. 277. J. G. Frayne and R. R. Scoville, “Analysis tronic mixer,” Jour. SMPE, 52: 676-683, 329. H. Fletcher, “Stereophonic reproduction and measurement of distortion in variable- June 1949. from film,” Jour. SMPE, 34: 606-613, June density recording,” Jour. SMPE, 32: 648- 304. A. C. Norwine, “Devices for controlling 1940; ibid., “Stereophonic sound-film sys- 673, June 1939. amplitude characteristics of telephonic sig- tem-general theory,” Jour. SMPE, 37: 278. Kurt Singer and Robert V. McKie, “Cross- nals,” BellSys. Tech. Jour., 77: 539-554, Oct. 339-352, Oct. 1941. (Seep. 352.) modulation compensator,” Jour. SMPTE, 1938. 330. J. C. Steinberg, “The stereophonic sound- 63: 77-81, Sept. 1954. 304a.D. E. Maxwell. “Dvnamic oerformance of ,I film qystem-pre- and post-equalization of 279. F. G. Albin, “A dynamic check on the proc- peak-limiting amplifiers,” Proc. I. R.E., 35: compandor systems,” Jour. SMPE, 37: 366- essing of film for sound records,” Jour. 1349-1356. Nov. 1947. 379, Oct. 1941. SMPE, 25: 161-170, Aug. 1935. 305. Kurt Singer, “Preselection of variable-gain 331. W. B. Snow and A. R. Soffel, “Electrical 280. C. E. Kenneth Mees, “The history of pro- tubes for compressors,” Jour. SMPE, 52: equipment for the stereophonic sound-film fessional black and white motion picture 684-689, June 1949. system,” Jour. SMPE, 37: 380-396, Oct. , films,” Jour. SMPTE, 63: 134-137, Oct. 306. Harold E. Haynes, “New principle for elec- 1941. I 1954. tronic volume compression,” Jour. SMPTE, 332 E. C. Wente, R. Biddulph, L. A. Elmer and 281. R. M. Corbin, N. L. Simmons and D. E. 58: 137-144, Feb. 1952. A. B. Anderson, “Mechanical and optical I Hyndman, “Two new Eastman fine-grain 307. W. C. Miller, “Volume control by the equipment for the stereophonic sound-film sound recording films,” lour. SMPE, 45: squeeze track,” Jour. SMPE, 75: 53-59, system,” Jour. SMPE, 37: 353-365, Oct. 265-285, Oct. 1945. July 1930. 1941. 282. Hollis W. Moyse, “DuPont fine-grain sound 308. E. H. Hansen and C. W. Faukner, “Me- 333. W. E. Garity and J. N. A. Hawkins, “Fanta- films-Types 232 and 236,” Jour. SMPE, chanical reversed-bias light-valve record- sound,” Jour. SMPE, 37: 127-146, Aug. 45: 285-293, Oct. 1945. ing,” Jour. SMPE, 26: 117-127, Feb. 1936. 1941. 283. L. A. Jones and Otto Sandvik, “Photo- 309. J. G. Frayne and H. C. Silent, “Push-pull 334. W. E. Garity and W. Jones, “Experiences in graphic Characteristics of sound recording recording with the light-valve,’’ Jour. road-showing Walt Disney’s Fantasia,” film,” Jour. SMPE, 74: 180-203, Feb. 1930. SMPE, 37: 46-64, July 1938. Jour. SMPE, 39: 6-15, July 1942. (Repetition of ref. 173.) 310. John G. Frayne, T. B. Cunningham and 335. E. H. Plumb, “Future of Fantasound,” 284. C. E. Ives and J. I. Crabtree, “Two new V. Pagliarulo, “An improved 200-mil push- Jour. SMPE, 39: 16-21, July 1942. films for duplicating work,” Jour. SMPE, pull density modulator,” Jour. SMPE, 47: 336. See group of papers on stereophonic sound, 29: 317-325, Sept. 1937. 494-518, Dec. 1946. Jour. SMPTE, 61: No. 3, Sept. 1953, Pt. 11. 285. Progress Report, Jour. SMPE, 48: 305, Apr. 31 1. Lawrence T. Sachtleben, “Wide-track op- 337. Panel Discussion, J. K. Hilliard, moderator 1947. tics for variable-area recorders,” Jour. “Equipment for stereophonic sound repro- 286. C. R. Daily, “Improvement in sound and SMPE, 52: 89-96, Jan. 1949. duction,” Jour. SMPTE, 62: 228-237, picture release through the use of fine-grain 312. J. E. Jenkins and S. E. Adair, “The control- Mar. 1954. film,” Jour. SMPE, 34: 12-25, Jan. 1940. frequency principle,” Jour. SMPE, 22: 193, 338. L. D. Grignon, “Experiment in stereo- 287. W. H. Offenhauser, Jr., “Notes on the ap- Mar. 1934. phonic sound,” Jour. SMPTE, 62: 64-70, plication of fine grain film to 16mm motion 313. J. G. Frayne and F. P. Herrnfeld, “Fre- Sept. 1953, Pt. 11. pic%ures,”Jour. SMPE, 41: 374-388, Nov. quency-modulated control-track for movie- 339. Progress Committee Report, Jour. SMPTE, 1943. tone prints,” Jour. SMPE, 38: 111-124, 62: 338, May 1954. 288. J. R. Wilkinson and F. L. Eich, “Labora- Feb. 1942. 340. Progress Committee Report, Jour. SMPTE, tory modification and procedure in connec- 314. G. I,. Dimmick and H. Belar, “An im- 60: 535, May 1953. See also Progress Com- tion with fine-grain release printing,” proved system for noiseless recording,” mittee Report, Jour. SMPTE, 64: 226-228, Jour. SMPE, 38: 56-65, Jan. 1942. Jour. SMPE, 23: 48-54, July 1934. May 1955. 289. Report on the adaptation of fine-grain films 315. D. J. Bloomberg and C. L. Lootens, “Class 341. Russell J. Tinkham, “Stereophonic tape , to variable-density sound technics, Jour. B push-pull recording for original nega- recording equipment,” Jour. SMPTE, 62: SMPE, 34: 3-11, Jan. 1940. tives,” Jour. SMPE, 33: 664-669, Dec. 71-74, Jan. 1954. 290. V. C. Shaner, “Note on the processing of 1939. 342. C. Robert Fine, “Perspecta Stereophonic c Eastman 1302 fine-grain release positive in 316. A. C. Blaney, “Notes on operating ex- Sound,” paper presented at 75th SMPTE P Hollywood,” Jour. SMPE, 38: 66-73, Jan. perience using the direct positive push-pull Convention, Washington, D. C., May 1954. 1942. method of recording,” Jour. SMPE, 42: 343. R. Gorisch and P. Gorlich, “Reproduction 291. E. W. Kellogg, “A comparison of variable- 279-282, May 1944. of color film sound records,” Jour. SMPE, density and variable-width systems,” Jour. 317. A. C. Blaney and G. M. Best, “Latest de- 43: 206-213, Sept. 1944. SMPE, 25: 203-226, Sept. 1935. velopments in variable-area processing,” 344. Alan M. Glover and Arnold R. Moore, 292. John G. Frayne, “Comparison of recording Jour. SMPE, 32: 237-245, Mar. 1939. “Phototube for dye image sound track,” processes,” (reprint) Jour. SMPTE, 59: 318. W. J. Albersheim, “Device for direct re- Jour. SMPE, 46: 379-386, May 1946. 313-318, Oct. 1952. production from variable-density sound 345. Lloyd T. Goldsmith, “Preliminary report of 293. N. Levinson, “New method of increasing negatives,” Jour. SMPE, 29: 274-280, Sept. research council photocell subcommittee,” the volume range of talking motion pic- 1937. Jour. SMPE, 48: 145-147, Feb. 1947. tures,” Jour. SMPE, 26: 111-116, Feb. 319. 0. L. Dupy, “A method of direct-positive 346. J. L. Forrest, “Metallic-salt track on Ansco 1936. variable-density recording with the light 16-mm color film,” Jour. SMPE, 53: 40-49, 294. G. L. Dimmick, “Optical control of wave- valve,” Jour. SMPTE, 59: 101-106, Aug. July 1949. shape and amplitude characteristics in vari- 1952. 347. Report of the SMPTE Color Committee, able-density recording,” Jour. SMPE, 33: 320. C. R. Keith and V. Pagliarulo, “Direct- L. T. Goldsmith, chairman, “Character- 650-663, Dec. 1939. positive variable-density recording with the istics of color film sound tracks,” Jour. 295. John G. Frayne, “Variable-area recording light valve,” Jour. SMPE, 52: 690-698, SMPTE, 54: 377, Mar 1950. with the light valve,” Jour. SMPE, 57: 501- June 1949. 348. Robert C. Lovick, “Optimum exposure of 520, Nov. 1948. 321. Lewis B. Browder, “Direct-positive variable- sound tracks on Kodachrome films,” Jour. 296. Lewis B. Browder, “Variable-area light- area recording with the light valve,” Jour. SMPTE, 59: 81-88, Aug. 1952. valve modulator,” Jour. SMPE, 57: 521- SMPE, 53: 149-158, Aug. 1949. 349. Robert C. Lovick, “Densitometry of silver 533, Nov. 1948. 322. L. I. Carey and Frank Moran, “Push-pull sulfide sound tracks,” Jour. SMPTE, 59: 297. E. M. Honan and C. R. Keith, “Recent direct-positive recording, an auxiliary to 89-93, Aug. 1952. developments in sound tracks,” Jour. magnetic recording,” Jour. SMPTE, 58: 350. C. H. Evans and J. F. Finkle “Sound track SMPE, 47: 127-135, Aug. 1943. 67-70, Jan. 1952. on Eastman color print film,” Jwr. SMPTE, 298. J. P. Livadary and S. J. Twining, “Vari- 323. John G. Frayne, “Electrical printing,” 57: 131-139, Aug. 1951. We-area release from variable-density Jour. SMPTE, 55: 590-604, Dec. 1950. 351. J. A. Maurer, “Present technical status of i I ‘ original sound tracks,” Jour. SMPE, 45: 324. C. N. Batsel and L. T. Sachtleben, “Some 16-mm sound-film, Jour. SMPE, 33: 315- 380-388, Nov. 1945. characteristics of 16mm sound by optical 325, Sept. 1939. 299. W. K. Grimwood, “Volume compressors for reduction and re-recording,’’ Jour. SMPE, 352. W. C. Miller, “Preliminary report of sound recording,” Jour. SMPE, 52: 49-76, 24: 95-101, Feb. 1935. Academy R Committee on Jan. 1949. 325. W. A. Mueller, “Audience noise as a limita- rerecording -mm release of 300. J, 0.Aalberg and J. G. Stewart, “Applica- tion to the permissible volume range of 35-mm feat PE, 45: 135- tion of non-linear volume characteristics to dialog in sound motion pictures,” Jour. 142, Aug. 1944. dialog recording,” Jour. SMPE, 37: 248- SMPE, 35: 48-58, July 1940. 353. E. E. Mastqwn, F, L. Putzrath and H E. .235, Sept. 1938. 326. H. I. Reiskind, “Multiple-speaker repro- Roys, “Miyll#& aound on 16-mm edge- 301. B. F. Miller, “Elimination of relative spec- ducing systems for motion pictures,” Jour. coated filq” Jaw. SMPTE, 57: 5%%66, tral energy distortion in electronic compres- SMPE, 37: 154-163, Aug. 1941. Dec. 1951. sors” fir. SMPE, 39: 317-323, Nov. 1942. 327. N. Levinson and L. T. Golds&& “Vita- 354. Geo. M. -, Thc Driue-In Tht&r* 302. M Mgerand K. Singer, “Factors gov- sound,” Jour. SMPE, 37: 147-153, Aug. Associated MI. Kansas City, Mo. Book err& befrequency response of a variable- 1941. Rev., Jour. SMPTE, 62: 401, May 1954.

456 355. R. H. Heacock, “Improved equipment for 372. Loren L. Ryder, “Motion picture studio use 389. W. E. Theisen, “The history of nitrocellu- drive-in theaters,” Jour. SMPTE, 60: 716- of magnetic recording,” Jour. SMPTE, 55: lose as a film base,” Jour. SMPE, 20: 259- 720, June 1953. 605-612, Dec. 1950. 262, Mar. 1933. 356. C. R. Underhill, Jr., “The trend in drive-in 373. Progress Report, Jour. SMPTE, 58: 400- 390. J. M. Calhoun, “The physical properties theaters,” Jour. SMPTE, 54: 161-170, 401, May 1952. and dimensional behavior of motion picture Feb. 1950. 374. John G. Frayne and John P. Livadary, film,” Jour. SMPE, 43: 227-267, Oct. 1944. 357. S. J. Begun, “Recent developments in the “Dual photomagnetic intermediate studio 391. Charles R. Fordyce, “Improved safety mo- field of magnetic recording,” Jour. SMPE, recording,” Jour. SMPTE, 59: 388-397, tion picture film support,” Jour. SMPE, 48: 1-13, Jan. 1947. Nov. 1952. 51: 331-350, Oct. 1948. 358. G. W. Carpenter and W. L. Carlson, US. 375. R. H. Ranger, “Sprocketless synchronous Pat. 1,640,881, filed Mar. 26, 1921; issued magnetic tape,” Jkr. SMPTE, 54: 328- 392. See Appendix. Aug. 30, 1927. 336, Mar. 1950. 393. Report of Historical Committee. Jour, 359. M. Camras, “Theoretical response from a 376. D. G. C. Hare and W. D. Fling, “Picture- SMPE, 17: 655-663, Oct. 1931. Address by magnetic wire record,” Proc. IRE and Waves synchronous magnetic tape recording,” Robert A. Millikan and communication by and Electrons, 34: No. 8, 597, Aug. 1946. Jour. SMPTE, 54: 554-566, May 1950. Wm. R. Ballard. 360. M. Camras, “Magnetic sound for motion 377. George Lewin, “Synchronous 1/4-in. mag- 394. Paul Schubert, The Electric Word, The Rise pictures,” Jour. SMPE, 48: 14-28, Jan. netic tape for motion picture production,” of Radto; The MacMillan Co., New York. 1947. Jour. SMPTE, 56: 664-671, June 1951. 1928, pp. 108 to 129. 361. Herr, B. F. Murphy and W. W. Wetzel, 378. Walter T. Selsted, “Synchronous recording K. 395. W. Rupert Maclaurin, Invention and Innova- “Some distinctive properties of magnetic- on 1/4- in. magnetic tape,” Jour. SMPTE, tion in the Radio Industry, The Macmillan recording media,” Jour. SMPE, 52: 77-88, 55: 279-284, Sept. 1950. Co., New York, 1949, pp. 95-98, 129. Jan. 1949. 379. Warren R. Isom, “Synchronized recordings 362. H. A. Howell, “Magnetic sound recording on perforated tape,” Jour. SMPTE, 63: 26- 396. Gleason L. Archer, History of Radio, publ. by on coated paper tape,” Jour. SMPE, 48: 28, July 1954. the Am. Hist. SOC.Inc., New York, 1938, 36-49, Jan. 1947. 380. Edward Schmidt, “Commercial experiences pp. 106-108,114. 363. Hershel Toomin and David Wildfeuer, “The with magnastripe,” Jour. SMPTE, 60: 463- 397. Irving Langmuir, “The effect of space mechanism of supersonic frequencies as ap- 469, Apr. 1953. Part 11. charge and residual gases on thermionic plied to magnetic recording,” Proc. IRE, 381. Andrew H. Persoon, “Magnetic striping of currents in high vacuum,” Phys. Rev., 32: 664-668, NOV.1944. photographic film by laminating process,” N.S. Vol. 11: No1 6, Dec. 1913. 364. Lynn C. Holmes and Donald L. Clark, Jour. SMPTE, 60: 485-490, Apr. 1953, Pt. 398. Irving Langmuir, “The pure electron dis- “Supersonic bias for magnetic recording,” 11. charge and its applications in radio teleg- Electronics, 18: 126, July 1945. 382. Thomas R. Dedell, “Magnetic sound raphy and telephony,” Proc. IRE, 3: 261, 365. G. L. Dimmick and S. W. Johnson, “Opti- tracks for processed 16-mm motion picture 1915. mum high-frequency bias in magnetic re- film,” Jour. SMPTE, 60: 491-500, Apr. 400. Just and Hanaman, U.S. Pat. 1,108,502, cording,” Jour. SMPE, 51: 489-500, Nov. 1953, Pt. 11. “Incandescent bodies for electric lamps,” 1948. 383. B. L. Kaspin, A. Roberts, Jr., Harry Rob- filed 1905-issued 1912. 366. Research Council Basic Sound Committee, bins, and R. L. Powers, “Magnetic striping (Acad. Mot. Pict. Arts and Sciences), Dis- techniques and characteristics,” Jour. 401. W. D. Coolidge, “Ductile tungsten,” Trans. cussion of Magnetic Recording, Jour. SMPTE, 60: 470-484, Apr. 1953, Pt. 11. AIEE, 2.9: Pt. 2, 961-965. SMPE, 48: 50-56, Jan. 1947. 384. Edward Schmidt and Ernest W. Franck, 402. C. G. Fink, “Ductile tungsten and molyb- 367. Earl Masterson, “35-mm magnetic record- “Manufacture of magnetic recording ma- denum,” Trans. Electrochcm. Soc., 17: 229, ing system,” Jour. SMPE, 51: 481-488, terials,’’ Jour. SMPTE, 60: 453-462, Apr. 1910. Nov. 1948. 1953, Pt. 11. 403. Irving Langmuir, “Thermal conduction 368. Dorothy O’Dea, “Magnetic recording for 385. Ernest W. Franck, “A study of dropouts in and convection in gases at extremely high the technician,” Jour. SMPE, 51: 468-480, magnetic film,” Jour. SMPTE, 60: 507-515, temperatures,” Trans. Electrochem. Soc., 20: Nov. 1948. Apr. 1953, Pt. 11. 225,1911. 369. 0. B. Gunby, “Portable magnetic recording 386. G. A. del Valle and L. W. Ferber, “Notes 404. Irving Langmuir, “Convection and conduc- system,” Jour. SMPE, 52: 613-618, June on wear of magnetic heads,” Jour. SMPTE, tion of heat in gases,” Phys. Rev., 34: 401, 1949. 60: 501-506, Apr. 1953, Pt. 11. 1912. 370. Progress Committee Report, Jour. SMPTE, 387. J. D. Bick, “Methods of measuring surface 56: 572-573, May 1951. induction of magnetic tape,” Jour. SMPTE, 405. Irving Langmuir, “Nitrogen filled lamps,” 371. C. C.Davis, J. G. Frayne and E. W. Temp- 60: 516-525, Apr. 1953, Pt. 11. Gen. Elec. Rev., 16: 688, 1913. lin, “Multichannel maanetic film recordinc 388. Ellis W. D’Arcy, “Standardization needs for 406. Irving Langmuir and J. A. Orange, “Tung- and reproducing unit,” Jour. SMPTE, 58 16 mm magnetic sound,” Jour. SMPTE, 60: sten lamps of high efficiency,” Trans. 105-118, Feb. 1952. 526-534, Apr. 1953, Pt. 11. AIEE, Pt. 11, 32: p. 1935, 1913.

Errata

In the second installment of this paper, published in the July Journal, the titles appearing under Figs. 8 and 9 should be interchanged; in the title to Fig. 14, “variation” should read “various”; and the title to Fig. 16 should read : “Horizontal cross section showing sound passages of folded low-frequency horn.”

Kellogg: History of Sound Motion Pictwrei 437