International projectionist (Jan-Dec 1939)

r °/vywv/vwyvV/V/v/\/ L 25.0 ^ O .5 1.0 1.5 SECONDS FIGURE 3 Chart showing objectionable flicker amounting to about 8 to 10% variation in transmission 24 0 FILM INCHES FIGURE 4 Flicker due to printing: transmission change caused by a ripple voltage of 10 to 15% Negative film is known to have random changes in sensitivity, and some stocks have cyclic changes occurring at at a rate of one cycle in 7 to 12 seconds, at a speed of 90 feet per minute. These variations are not in themselves too serious, but should they fall in phase with other cyclic changes, then the resulting flicker would be noticeable. Stocks having cyclic and random sensitivity have been submitted for use. In general these defects have been minimized. • Irregular Camera Motion Irregular camera motion is one of the worst offenders at present, but the motor system is not blameless and can be the cause of flicker. For some time the interlock motor system, when used, was condemned for all this trouble; the fact is that the interlock motor system was not the source, but its basis of operation allowed the trouble to persist and frequently amplified it. It would appear that, to obtain a steady exposure, the speed of the rotating shutter, which exposes the film, should be as smooth and constant as the movement of film through a sound recorder. A great deal of time and money has been spent by sound equipment manufacturers and users to reduce flutter, and, as previously mentioned, picture flicker is nothing more than flutter. However, cameras generally use a slipping belt directly coupled to the shutter shaft for a film take-up mechanism. Belt condition greatly influences the steadiness of take-up, and ■each instant that a sudden change in load occurs the motor system reflects that change. Even with a motor having no resilience, changes in load can cause dicker. The camera undoubtedly contains mechanical inductances and capacities (which would include the shutter, motor rotors, gears, backlash, motor air-gap, flux, etc.) that can become resonant. Even though these reactances -are inherently stable, the system might be thrown into oscillation by a sudden shock of small magnitude. This is evidenced by circumstances that have occurred when belt condition, mechanical looseness, and shutter action have all combined to become oscillatory and persistent at a rate well within the greatest disturbing region of Fig. 1. It was while working on a new motor system in conjunction with Erpi that a full realization of the true nature of the difficulty was reached. A few clutches have been tried that gave varying degrees of improvement but none completely solved the problem. Fig. 2 shows a chart of a section of picture negative in which the flicker was just perceptible, representing about 3% variation in transmission. Fig. 3 shows a similar chart having flicker amounting to about 8% to 10% variation in transmission. Those who have never seen the action of a camera shutter might observe the opening or closing edge of the shutter with a stroboscope, which is accurately synchronized with the motor, as either of these edges pass the aperture. Obviously, any variation in the shutter while it is fully open will have no deleterious effects. Development of negative is suspected of causing some variations, but no conclusive data are yet available. It should not be deduced, however, that the laboratory is entirely faultless. (B) Printing. — When the printer light is supplied from a generator or rectified alternating current, sufficient filtering must be used to reduce intensity changes to a small value. The ripple voltage should be less than 1 per cent. It is true that the normal ripple frequencies are beyond the greatest disturbing flicker rate, but the existence of 120 cycles in conjunction with other flicker frequencies produces a creeping density pattern in the projected picture. Frequencies increasingly higher than 120 cycles would undoubtedly cause less and less trouble. This particular effect is greatly dependent upon the amount of light and the density of the various parts of the scene. It becomes most apparent in scenes including dark skies such as in night shots made with filters in the daytime. Fig. 4 shows the transmission change caused by a ripple voltage of 10 to 15 per cent. Periodic flicker has been traced to printer motion on an earlier type of machine, but no data exist on machines of current manufacture. In the case under investigation the flicker was caused by the belt splice which created created periodic film-speed changes as the stock passed the aperture; i.e., as the belt splice passed over the pulley the effective radius of the pulley was changed, causing a corresponding filmspeed variation. Fig. 5 shows this effect under two conditions, that of normal FIGURE 5 Flicker effect in printing, due to belt splices 12 3 4 BELT REVOLUTIONS SEPTEMBER 1939 19