How and Why the Talkies
By HORATIO W. LAMSON, Engineering Department
Part Two
In the December issue of the EX-I'EHIMENTER we outlined the general problems of the talking movies and discussed in some detail the wax disc method of recording and reproducing synchronized sound. We now propose to consider the optical method whereby a photographic record, corresponding to the fluctuating sound impulses, is obtained upon a film in the studio and subsequently reproduced as sound in the theater.
There are two fundamental forms which such a film record may take. The one most commonly employed has a constant transverse width and an intensity or "density" which varies from point to point along the film in accordance with the frequency and amplitude fluctuations of the corresponding sound waves. Such a record is shaded in appearance and similar in character to a photograph of a heavily banded light spectrum. It may be obtained by varying the intensity of the light source, either directly or indirectly, or by changing the effective width of the narrow slit opening through which the film is exposed. In either case the density across the record is constant at any given point along the film, giving thus the characteristic banded effect.
On the other hand, in the second type of record, a constant source of illumination is employed, while the electrical impulses corresponding to the sound waves operate a mechanism that serves to vary the relative amount of the transverse slit which is illuminated at any given instant. This produces a black-and-white non-shaded record having a fine sawtooth appearance.
There are three distinct methods of optical recording which may be described here, all of which are fun damental^- adapted to produce a banded type of record.
The first makes use of the neon or similar type of glow lamp which is so well known in the art of television. The intensity of the light, emitted by such a lamp can be varied rapidly and easily by a fluctuating voltage applied to it. If, now, our film is driven uniformly along behind a narrow transverse slit which is illuminated on the opposite side by such a lamp, we have the means of producing a banded film record corresponding to the variations of the sound waves picked up by the studio microphones.
The second method utilizes a constant source of light and employs an ingenious device known as a "light-valve." This consists of two parallel duraluniinum tapes each six mils wide and three mils thick. These are so placed in the optical system that, when at rest, they form an effective slit which, viewed against the source of light, presents an opening two mils wide by one quarter of an inch long. By means of a high grade optical system an image of this slit in the light valve is thrown onto the the film in the form of a transverse line of light one-eighth of an inch long and, normally, only one mil wide.
The two duraluminum tapes form an electrical loop circuit and they are so located in a steady magnetic field that, when a pulsating current is passed through them, they move in opposite directions. In this manner the effective gap opening between the tapes, and, hence, the width of the image line on the film, is varied according to the frequency and amplitude of the electrical impulses. Such a modulation of the light gives, of course, the characteristic banded record on the moving film. We note that in this case the time during which each spot on the film is expose«! to a constant light source varies, while in the other two methods described eacli spot on the film record is illuminated for the same time interval by a modulated light intensity.
It is found desirable in practice to adjust the tension on the tapes until they have a natural frequency of about 7000 cycles per second. Under this condition a 100 per cent, modulation of the light, i.e., opening the valve slit to a maximum of four mils and just closing it completely requires about ten milliwatts of power at the lower audio frequencies and about 0.1 milliwatt at the natural frequency of the tapes.
A third method of optical recording utilizes an interesting device known as the Kerr cell. A beam of light of constant intensity is passed first through a Nicol's prism which polarizes the beam in a particular plane. It is then passed through a narrow gap between two electrodes and subsequently through a second Nicol's prism set at 45 degrees to the first. The gap between the electrodes in the Kerr cell is filled with nitro-benzol, a liquid which has the property of rotating the plane of polarized light passing through it when subjected to an electrostatic field to a degree proportional to the impressed voltage. Obviously, then, a modulation of the effective intensity of the light source may be produced by applying an alternating potential to the electrodes, so that, if we employ a fixed transverse slit against the film, a banded record will result. The separation of the electrodes and the length of the light path between them determine the voltage necessary to produce 100 per cent, modulation, i.e., variation
between full transmission and total extinction of the light. As in the case of the neon lamp, the degree of modulation of the Kerr cell is essentially independent of frequency, but in the Kerr cell the degree of modulation is proportional to a cosine function of the amplitude which, obviously, limits the usefulness of this device.
In all three methods of optical recording the nominal effective time of exposure is about 1/18,000 second, corresponding to the nominal film speed of ninety feet per minute. This means that modulation of the record at a frequency of 18,000 cycles or higher would be nil with an increasingly better modulation as we go below 18,000 cycles. In the workable audio range, however, the modulation is satisfactory, or can be made so by the use of equalizers.
The technique of studio recording follows along the general lines previously described. The responses of the several studio microphones, properly mixed, are amplified sufficiently to operate whichever type of recording device is used. In general two separate film records are made, one for the sound and one for the photography. This permits a different technique of development for the two films, which is very desirable. Synchronization is accomplished by an interlocking electrical drive system, which contains mechanical filters in the drive of the sound film to maintain a constant and uniform rate of travel past the exposure slit.
It is customary in both disc or photographic recording to make two identical sound records and subsequently to choose the better for printing the released positive films or preparing the playing records. The optical sound record, as printed on the films sent out to the theaters, takes the form of a strip about one-eighth of an inch wide along one side of the picture. The picture and the sound strip are printed separately on the positive film, the space occupied by one being shielded from the light while printing the other.
Great care must be taken in selecting both the positive and negative raw film stock to be used in sound picture work. Any irregularities in the transparency of the film or emulsion will, of course, generate unwelcome "background" noises in the final reproduction. While the eye can barely detect a two per cent, change in film density sudden irregularities of only one-tenth of one per cent, will give rise to an audible background noise.
In reproducing the optical sound record in the theater we employ another device used in the art of television, namely, the photo-electric cell. When subjected to a steady polarizing voltage the photo-electric cell allows a current to How through it which is proportional to the intensity of light falling upon the cell. A narrow transverse slit is interposed between the cell and a constant source of light. This slit extends across the portion of the film carrying the sound record so that the intensity of the light passing at any instant into the photo-electric cell depends either upon the relative density of the banded film back of the slit or upon the width of the "cut-off" portion of the saw-tooth film record at the point in question.
Thus if the film is drawn uniformly at the original speed across the slit we will obtain a pulsating current in the photo-electric cell circuit which will be a reproduction of the current in the studio microphones.
When projecting a motion picture the film is advanced intermittently at the rate of sixteen "frames" per second, each frame being stationary for the brief instant during which light is passing through it to the screen. Obviously, then, the synchronized picture and sound record can not be adjacent on the film. In practice they are spaced about fifteen inches apart along the film, thereby allowing for a "loop" to take up the intermittent slack between them. Mechanical filters are used in the drive to insure an extreme uniformity of motion past the photo-electric cell slit.
The electrical impulses obtained from the photo-electric cell are extremely small in amplitude. A two-stage resistance-coupled amplifier is ordinarily necessary to bring them up to an energy level comparable with that obtained directly from the electromagnetic pickup used in the disc method of reproducing. On ac count of the high impedance of the photo-electric cell circuit, it is desirable to build this amplifier into the same container which holds the photo-electric cell. The output of this amplifier, at low impedance, is then carried to the fader and, from this point on, the same speech am-plif3'ing system used with the disc method is employed, likewise the same technique of fader operation, monitoring, and control of the output panel.
A large proportion of the theaters in which synchronized sound installations have been made are equipped with machines designed for projecting either the disc or the film type of record. They can, therefore, utilize the standardized product of all producers.
A few statistics may be of interest in closing. A number of concerns are now developing and placing on the market equipment for recording and reproducing sound pictures. Among them are: the Electrical He-search Products, Inc., a subsidiary of the Western .Electric Company, who use both the disc record (Vita-phone) and the film record (Movietone) ; the 11. C. A. Photophone (a film record) ; and the General Talking Pictures Corporation (De Forest Phonofilm), likewise a film record.
Among the producers using the E. R. P. I. system are:
1. Paramount Famous Players
Easky Corp.
3. Warner Brothers.
7. Christie Comedies.
8. Victor.
The following producers are licensed under the R. C. A. Photo-phone system :
Tiffany-Stahl.
Pathe.
Mack Sennett. Educational.
In conclusion we wish to acknowledge our indebtedness to the Bell Laboratories Record for certain data used in this article.
Qfie General Radio
Post a comment