American cinematographer (Feb-Dec 1929)

Six AMERICAN CINEM ATOGRAPHER February, 1929 Recording with an Oscillograph : Variable Width Records The difficulty of producing high quality sound records of the variable-density type has already been emphasized in a previous paper," where it was shown that the characteristics of photographic materials are such that improper Fig. 10 Optical system for making variable width sound record with an oscillograph. exposure or development will introduce distortion. This source of distortion can be eliminated by the use of sufficient care in the photographic phase of the problem, but it is doubtful whether this is justified at the present time. The variable width type of record is inherently free from this type of distortion and is easily produced by an optical system such as that shown in Fig. 10. In this arrangement the source of light is focused on the oscillograph mirror by means of a spherical lens. As before, the mirror is drawn as though it produced no deflection of the beam. In the vertical plane, shown in the lower diagram, the mirror is refocused on the close-up slit by means of a cylindrical lens. In the horizontal plane the beam is allowed to diverge and cover the width of the sound track. A stop at the condensing lens C covers half of the area of the lens. This stop is reimaged on the film by means of the single lens close to the mirror, this lens being again shown double in the diagram since the light traverses it twice. The neutral position of the mirror is adjusted so that the image of the edge of the diaphragm at lens C occupies the center of the sound track. As the signal is impressed on the oscillograph, the image of the diaphragm moves from side to side and produces records of the well-known variable width type. Since the light-gathering properties of the oscillograph have already been discussed at some length in the previous paper," it is only necessary to refer here briefly to the general conclusions. As in the previous case, shown in Fig. 7, the illumination is proportional to the product of the angles 0, and ©2. The angle ©2 may have a value equal to that of 92, in Fig 7. On the other hand, the value of ©, depends on the distance from the mirror to the film which, in turn, depends on the sensitivity of the oscillograph vibrator. It is usually possible to obtain a deflection of one inch at a distance of ten inches without heating the vibrator enough to give it an undesirable frequency response. With a sound track 0.1 inch in width, therefore, the proper place for the mirror is one inch from the film. If it is placed closer than this, the full width of the sound track will not be utilized without overloading the vibrator. If it is placed at a greater distance, Qi, the illumination will be reduced proportionately. It is frequently impracticable to place the vibrator as close to the film as required by the optical considerations. In such cases, as we have seen before, the mirror may be imaged at the proper point, perferably by means of a cylindrical lens inserted in the upper diagram. Such a lens, however, is powerless to increase the illumination over that provided by the scheme shown in Fig. 10. f. "The Optical System of the Oscillograph and Similar Recording Instruments," by A. C. Hardy, J.O.S.A., R.S.I., 14, June, 1927, p. 505. 7 "The Rendering of Tone Values in the Photographic Recording of Sound," by A. C. Hardy, Trans. Soc. M. P. Eng., Vol. XI, No. 31, p. 475, (1927). With a mirror 0.5 mm. in width and 25 mm. from the film, the numerical aperture in the upper diagram is only 0.01. Since this system, however, requires no modulation control with its attendant reduction in exposure it is very satisfactory for recording even on cine positive film. This system has been considered with reference to the close-up slit arrangement only. From previous consider ations, however, it is obvious that the slit could be imaged on the film, and it is likewise apparent that such a procedure is justified only as a matter of convenience from the mechanical standpoint rather than from any superior light-gathering characteristics of the imaged slit system. Reproducing We have already found that the illumination of the reproducing slit should be 200 lumens per square centimeter. This corresponds to the placing of a 200 candlepower lamp at a distance of 1 centimeter or the equiv Fl 1 1 Photoelectric Cell Diagram of optical system of photoelectric cell reproducer. alent. The design of incandescent lamps is such that the bulb usually prevents placing the filament sufficiently close to the reproducing slit. As a consequence, it is customary to employ an optical system similar to that shown in Fig. 11. The filament is here imaged on the reproducing slit and the illumination may be calculated by means of equation 4. Since the brightness of the average tungsten filament operated in such a way as to insure a reasonable life is of the order of 1200 candles per square centimeter, a microscope objective of 0.25 numerical aperture will produce an illumination at the slit of 235 lumens per square centimeter. As we have already seen, this will provide an average photoelectric cell current of about a microampere. The objections to a close-up slit apply even more forcibly in reproducing than in recording. Consequently it is preferable to use the system shown in Fig. 12, wherein Phototlecmc Cell Diagram of reproducer optical system with close-up slit eliminated. the filament is imaged on the microscope objective O by a condensing lens C and a slit of generous dimensions is imaged on the film. If the objective O is filled with light, the flux reaching the photo-electric cell will be the same in this case as with the system shown in Fig. 11. Conclusion The declared purpose of this paper was to illustrate the general method of computing the amount of light which passes through any optical system. Although this general method has been freely illustrated by examples chosen from among the better known methods of sound recording, it is impossible to compare the merits of the various systems without assigning actual numerical values to the quantities involved. This is due to the fact that the amount of light available with most systems depends upon mechanical or electrical limitations rather than optical ones. For example, in the case of a string galvanometer the amount of exposure on the film with a properly designed optical system depends on the displacement which may be given to the string and its current capacity. It is hoped, therefore, that the numerical values herein presented for purposes of illustration will not be interpreted as a final comparison of the merits of the various systems of recording and reproducing. The question of image perfection has been entirely ignored in the present paper for lack of space. Charles Stumar, A. S. C, writes from Arosa, Switzerland, that he is engaged in shooting Universal's first German production, directed by William Dieterk, and starring Lien Deyers. The title of the picture is "The Triumph of Life." Charlie will be absent several months.