Radio Broadcast (May 1929-Apr 1930)

TECHNICAL DATA SOUND MOTION PICTURES BY CARL DREHER Magnitudes in Reproduction THE principal job of the engineer in any field is to know quantitative relations. The purpose of this article is to discuss the usual light intensities, energy levels, and degree of amplification in sound motion picture reproducing systems. Some of this material has been presented from one viewpoint in the February issue of this department under the heading of "Further Data on Photo-Cell Characteristics." Much of the additional data given below is from the paper by Arthur C. Hardy, "The Optics of Sound Recording Systems," published in Vol. XII, Number 35, of the Transactions of the Society of Motion Picture Engineers. Given a point source of light (one which is relatively small compared to other dimensions in the optical system) the intensity in a given direction depends on the amount of radiation from the source per unit of solid angle in that direction. If the source radiates L lumens in the solid angle w, the intensity, /, is given by I = (l) When the light from a point source of intensity, /, falls on a surface placed at right angles at a distance cZ, the illumination is given by (2) The above equations differ in some respects from their analogues in radio transmission, but only because of practical differences resulting from the order of dimensions in the two fields and the technique used in transmission, pick-up, and measurement. The lumen, the unit of light radiation or flux, is analogous to the watt, but it is a narrow unit, applicable only to radiation of a certain kind, i.e., visible radiation, while the watt is a general energy unit. In the region of maximum visibility a lumen corresponds to about 1.5 milliwatts. Light intensity as expressed by (1) above is expressed in terms of energy in a given solid angle, while radio field intensity is expressed as a potential (volts per meter) because of the method of pick-up used, involving an antenna. The unit of light intensity is the candle, corresponding to one lumen per unit of solid angle. Illumination may then be expressed in foot-candles, meter-candles, or centimetercandles. The last, as may be noted from a further inspection of (1) and (2), is the same as lumens per square centimeter. The latter is a good practical unit to keep in mind. If the source of light is too large to be considered as a point, the concept of brightness must be introduced. Brightness may be defined as intensity over a given area. It is, therefore, measured in candles per square centimeter. Since one candle is, by definition, Source nn n n n ,6 n o q i (A) Correct splice, quiet Fig. 3. a radiation of one lumen per unit solid angle, it follows that a surface of unit brightness, as an incandescent lamp filament, sends out in a given direction one lumen per unit solid angle for each square centimeter of its area. Professor Hardy, in the paper cited, sums up the photometric units in the following table: Table of Photometric Units QUANTITY Flux Intensity Illumination Brightness UNIT Lumen Candle Lumen per square centimeter or centimeter candle Candle per square centimeter "■ Window In reproduction from film the track is run at a constant speed past a thin rectangle of light, which shines through the film onto the window of a photo-electric cell. In its simplest form the mechanism may consist of a slit about 0.001 inch in width close up against the film and brightly illuminated. Such a system is shown in Fig. 1. The light rectangle on the film must be narrow compared to the highest frequency it is proposed to reproduce, so that 6000 cycles, for example, with one cycle occupying 0.003 inch on the track, requires an 0. 001-inch slit close up against the film (say the same distance) to prevent the light from spreading out between the slit and the film. The area of such a slit is roughly 0.0005 square centimeter (0.001 inch in the direction of travel of the film by 0.1 inch transversely, corresponding to 0.00254 cm. by 0.254 cm.) A good commercially available photo-cell may have a sensitivity of 10 microamperes per lumen. (See " Sound Motion Pictures, " February, 1929, Radio Broadcast, page 244). It is not desirable to operate with a photo-cell output of less than 1.0 microampere, because of noise interference considerations. It follows that we have to get at least 0.1 lumen of light into the cell. This necessitates illuminating a slit with the above dimensions with not less than 200 lumens per square centimeter, which, as the lumen per square centimeter corresponds to the centimeter candle, may be secured from a 200 candle-power lamp at a distance of one centimeter. Because the bulb of the lamp would be in the way if this were attempted, the lens system shown in Fig. 1 is employed. By this means the lamp may be moved away a convenient distance and the light focussed on the slit. It is found that the filament must be operated at a brightness of 1200 candles per square centimeter to meet these conditions. This is about the limit, in practice, at which a reasonable life may be expected. The close-up slit shown in Fig. 1 is not practicable for actual use in theatres, since an opening 0.001 inch wide up against a rapidly moving film cannot be kept free from foreign matter for any length of time. The optical expedient shown in Fig. 2, which, like Fig. 1, is copied from Hardy's paper, is accordingly adopted. The slit is moved away from its exposed position near the film, sealed within a lens system, made considerably wider and longer (say 0.004 by 0.4 inches), and then optically reduced to a light rectangle 0.001 by 0.1 inches on the film. With proper design of the optical elements the results are equivalent to the close-up slit system. With the light levels assumed above the photo-cell receives 0.1 lumen and, with the stipulated sensitivity of 10 microamperes per lumen, will yield an output of 1.0 microampere through the anode resistance of 2 megohms, corresponding to an output of 2.0 microwatts. If this is amplified by 70 db, or 10,000,000 times in energy, the amplifier output in, let us hope, undistorted audio energy, will be 20 watts. This output is, in fact, required to fill the average theatre of about 2000 seats. Assuming that the loud speakers have an efficiency of 20 per cent., we get four watts of sound energy out into the house after all the transformations of the system. Splicing Sound Films A VALUABLE article of a practical nature appearing in the January 5 Movietone Bulletin concerns the method of making a splice in a sound-movie film. If the splice is not properly treated a lamentably loud thump startles the audience in the theatre, as a result of the electrical impulse sent into the amplifier by the discontinuity in the sound track. Fig. 3, reproduced here from the drawing in the Movietone Bulletin, shows how this is obviated. The light is gradually shut off by a triangle painted with black or red lacquer (India ink is sometimes used) onto the sound track. The base of the wedge should be about § inch long along the line of sprocket holes, while the apex is at the splice on the inner edge of the sound track. If the splice is covered by too short a triangle the change in light intensity will remain abrupt and more or less noise in reproduction will result. If too long a triangle is painted, on the other hand, enough of the track may be obliterated to cause an interruption in the record. Some care is, therefore, necessary. The painting is done on the celluloid side of the film. Source, Fig. 2. • may, 1929 page 33 •