American cinematographer (Feb-Dec 1929)

Four AMERICAN CINEMATOGRAPHER February, 1929 The Optics of Sound Recording Systems By Arthur C. Hardy Associate Professor of Optics and Photography, Massachusetts Institute of Technology, Cambridge, Mass. Read at the Fall Convention of the S. M. P. E., 1928. Held at Lake Placid, New York. (Continued from January Cinematographer) Recording ivith a Kerr Cell In the systems that have just been considered, the Fig 4 brightness of the light source was modulated by introducing the amplified sound current into its supply circuit. Unfortunately there are only a few discharge tubes which will follow the rapid changes in sound intensity and the brightness of these sources is usually low. Until glow lamps of higher intrinsic brightness are developed, it will be necessary to use other sources of light for high quality recording since the amount of light available with the best of glow lamps at the present time is barely sufficient to expose cine negative film. The low resolving power and bad graininess of this material make it undesirable for sound recording purposes, and, as the desirable pho Fig 5 tographic characteristics are obtainable only at the sacrifice of sensitivity, it is customary to use a light of higher brightness with some sort of modulating device or light valve. The simplest of these is the well-known Kerr cell. The Kerr cell consists of a pair of electrodes in a liquid such as trinitrobenzene which has the property of becoming doubly refracting in an electrostatic field. A Nicol I Nicol Kerr Cell Fig. 6 Diagram of the Kerr cell recorder. sound recording system using such a cell is shown in Fig. 6. Light from any source of constant brightness, such as the filament of an incandescent lamp, passes through the collimating lens and then through a polarizing Nicol prism before entering the Kerr cell. On leaving the cell the light passes again through a Nicol prism and a lens which forms an image of the filament source on the slit in front of the film. The second Nicol prism is so adjusted with respect to the first that when no voltage is applied to the electrodes of the cell they transmit one quarter of the incident light. When the signal voltage is applied to the cell* the exposure of the film is increased or decreased in a manner which corresponds roughly to the original variations in sound intensity. 4 Since the numerical aperture of a microscope objective is usually known, it is frequently more convenient to use equa tion 4 instead of equation 6. The numerical aperture of a microscope objective used in air is equal to the sine of the angle 0 when the objective is used at the magnification for which it was designed. As the objective shown in Fig. 4 is used for reduction rather than for magnification, the sine of the angle 0 is equal to the indicated numerical aperture when the distance from S to O is roughly 160 mm. The sine of 0 varies inversely as the magnification plus one. Consequently, when the magnification is less than 1, as in Fig. 4, the illumination depends very little upon the magnification. For high values of the magnification it is nearly inversely proportional to the square of the magnification. 5 This assumes the use of a glow lamp of the same brightness and the same effective slit at the same magnification in both cases as explained in the last footnote. It is obvious that the optical system of Fig. 6 is fundamentally the same as that shown in Fig. 2 provided that the Nicol prisms and Kerr cell are of sufficient size. In other words if the second lens is the aperture stop, as Fig. 6 shows, the angle 6 may be made sensibly the same as in the case shown in Fig. 4. Although the normal setting of the Nicol prisms decreases the average illumination to one quarter of the amount available without them, the brightness of the filament of an incandescent lamp is many times greater than the brightness of any glow lamp having a reasonable life. Consequently, recording may be done with the Kerr cell arrangement on a less sensitive and better adapted type of photographic material. Recording with a Strina Galvanometer : String Vertical Of the mechanical devices which might be used for light valves, the simplest is undoubtedly the string galvanometer. This consists of a wire or ribbon called the string, suspended in a strong magnetic field. The passage of current through this string causes a deflection which may be made to open or close a suitable aperture. The string is normally adjusted to fill half of the opening when it carries no current. Then when a signal is impressed on the galvanometer, the resulting oscillations of the string cause corresponding variations in light flux which may be used for photographic recording. Ordinarily the film travels in a vertical plane so that the recording slit is horizontal. The string can be placed either vertically (across the slit) or horizontally (parallel to it). Sound records of the variable width type can be made with the string vertical, but the illumination is so low that such a procedure is not practical. To secure good quality, the natural period of the string must be less than that of the highest pitch to be recorded. This means that the string must be light and consequently incapable of large deflections. In practice it is difficult to obtain deflections greater than 0.001 inch in a galvanometer having a satisfactory frequency response curve. As this deflection must be magnified nearly 100 times to fill a sound track of the usual width, the available illumination is necessarily low (see footnote 4). By means of a suitable system of cylindrical lenses, the magnification between the string and the film may be 100 to 1 in the horizontal plane and 1 to 1 or better in the vertical plane. In the Kerr cell arrangement of Fig. 6 the magnification between the source and the film was 1 to 1 in both planes. Consequently, neglecting for the moment the absorption of light by the Kerr cell and the loss caused by the Nicol prisms, this type of string galvanometer arrangement can be only 1/100 as efficient in light-gathering power as the Kerr cell arrangement. If we continue to neglect the absorption in the cell but remember that the Nicol prisms reduce the illumination to one quarter of the value it would have otherwise, we find