US1995976A - Process for the production of sound films according to the amplitude process - Google Patents

Description

TUDE PR ESS Mar h 3 E. GERLACH PROCESS FOR THE PRODUCTION OF SOUND FILMS ACCORDING TO THE AMPLI Filed Nov. 4, 1930 V IVVENTOR IRWM/ER EM r I ATTofiM x/s i Patented Mar. 26, 1935 PROCESS FOR THE PRODUCTION OF SOUND FILMS ACCORDING TO THE AMPLITUDE PROCESS Erwin Gerlach, Berlin-Siemensstadt, Germany,

assignor. to Siemens & Halske, Aktiengesellschaft, Siemensstadt, near Berlin, Germany, a corporation oi Germany Application November 4, 1930, Serial No. 493,288 In Germany November 11, 1929 2 Claims.

The invention relates to a p'rocess'for the production of sound films which are made according to the amplitude process in which, as is well known, the width of the blackened portion of the film (such blackened portion being of uniform shade or density) varies transversely to the longitudinal direction of the film, proportionately to the sound variations to be recorded. In the reproduction of such sound films, there are produced in the case of certain sounds, in particular with the pronunciation of the S in human speech,

disturbing subsidiary sounds of medium or low frequency of clearly audible amplitude, the avoidance of which is the object of the present invention.

In order to explain the theory and practice of my invention, I have annexed hereto a drawing, in which: I

Fig. 1 illustrates diagrammatically the latent image which is present on an exposed but undeveloped variable width sound film.

Fig. 2 is a graphic diagram representing the intensity of exposure, the light slit, and the latent image.

Fig. 3 illustrates diagrammatically the developedvariable width sound film produced according to the present invention.

In the recording of the sound picture, a narrow beam of light is used, the dimension of which transversely to the movement of the film varies with the amplitude of the sound vibrations to be recorded, and which in this way produces on the traveling film strip. a curve varying with the" amplitude and frequency of the sound vibrations, said curve constitutin, the boundary between the exposed and the unexposed portions of the film. In general, the'beam of light is made as narrow as posslblepbut inthis optical and other difficulties are found which have hitherto not permitted a reduction of the effective width of the slit belowmcertain limit. The production of the disturbing subsidiary sounds is due to the impossibility of reducing the width of the beam of light below saidlimit.

If, for example, with a beam of light having a width which is approximately one-quarter of the wave length .of the vibrationto be recorded, it is attempted to record a pure sine wave appearing as the boundary between black and white areas then, instead of the pure sine form, there is obtained a distorted curve such as shown in Fig. 1.

On the other hand, the appreciable flattening at. the bottom or trough of the second half period of a vibration is of importance. The value of this flattening d is dependent on several factors, that (Cl. 274-4L6) is, on the amplitude of the vibration, thewidth of the beam of light, the speed of the film, and the frequency (in cycles) to be recorded. Assuming, for example, a film speed of about 500 millimeters per second and an effective light beam width of minimum practicable width, with a frequency of 8000 cycles, which is the one to be taken into account in the case of human sibilants, calculation indicates losses of about 10% in the total amplitude of the recording alternating current. The flattening of the amplitude at the trough means that in the reproduction, an effect is produced as if there were superposed on the signal alternating current, a direct current the amplitude g of which is still greater, in consequence of the simultaneous narrowing of the curve trough, than the value (1 of the flattening shown in the drawing.

As long as the amplitude of the'frequencies being recorded fluctuates but slowly asfor example in the case of musical sounds, the amplitude of the direct current also fluctuates slowly and no audible vibrations will result from these slow fluctuations. It is, however, different in the case where the amplitude of the vibrations being recorded is modulated rapidly and within wide limits as is, for example, the case with the human S, the frequency band of which (about 8000 cycles) is modulated in this way. The superposed direct current fluctuates corresponding to the rapid variation'of the amplitude, in a frequency which, however, lies within the limit of audibility, and the vibration values of which are rather large.

In the amplitude recording process, however, a further effect takes place which likewise has received no attention hitherto. If the slit forming the'beam of light is assluned as split up into a very large number (n) of narrow strips, then each of these strips will permit the passage of l/nth of the entire quantity of light transmitted through the entire slit. As will appear more clearly from the description to follow, these strips increase in height, and therefore in the quantity or intensity of light transmitted by each strip to the film, simultaneously. If the beam of light moves on a sine curve, each portion of every strip will also move on a' sine curve However, as the strips are laterallyspaced, the sine curves followed by the difierentstr'ips will be the sanie in size and shape but laterally spaced or shiftedrin phase. This will produce a central section which is a part of each sine curve (the black part of Fig. 1)

curves. 'Any usual method of producing variable width negatives of sound records will give a latent image having this gradually lightening fringe around a fully exposed core or base. These laterally spaced sine curves, taken together, form the strongly distorted curve shown in Fig. 1 and described above. If oaing to sudden variations in the amplitude of the sound being recorded, there should be a sudden change in the light effecting the exposure, the fact that the slit has a definite width will, as it were, cause the time of exposure corresponding to the additional exposure to be, for the first strip of the slit, l/nth of the maximum intensity of the light passing through the slit, and 2/nth for the second strip. Consequently, there is produced on the film, as shown by the lower part of Fig. 2, a latent image it the nature of a field having the same width as the slit b; within this field there is a shading increasing in depth gradually, so as to give the impression of a grey fiel' d. If the surge of light continues for a certain time, this grey field will become darker, until it is blackened to a degree corresponding to the intensity of the light passing through the slit. .In Fig. 2, the line a represents the intensity of light or height of the beam as it varies during the movement of the film from right to left, the time at being marked along the abscissa strips or slit sections, of which the first and second axis, while the ordinates represent the intensity of the light. The slit b through which the beam of light passes is represented as subdivided by (imaginary) vertical lines into n narrow vertical 40 the film there is a change in the intensity of exposure, that is, if the height of the beam suddenly increases as s represented by the right hand portion of line a, the light will strike that portion of the film above the dark lower section of c and will produce a latent image of the type represented by the upper part of c. Since the film is moving from right to left, that portion of the film which lies under the slit section 1 at the extreme left 'of the slit "b, will receive an additional exposure during only l/nth of the time which a (vertical) strip element of the film requires to cross the entire width s of the slit 1). The film strip lying under the second slit section from the left, 2, has, like thefilm strips under the other slit sections, received at the end of this period an exposure the length of which is l/nth. Therefore, by the time this second film strip disappears under the lefthand edge of the slit b, it has received an additional exposure during another period of l/nth,

so that its total additional exposure amounts to 2/nths of the time which a strip element of the film requires to cross the entire width of the slit b. Similarly, that strip element of the film which lies under the third slit section (counting from the left), will receive a total additional exposure mounting to 3/nths of the time mentioned above. The last strip element of the film (at the right hand edge of the slit b) will receive the greatest amount of additional exposure, and it will be obvious that from that moment on, the film will receive a uniform exposure, by the entire amount of light passing through the slit b. It will be noted that in Fig. 2 the vertical lines in the field c are progressively closer together toward the right, the increasing depth of this shading indicating (s being the width of the slit b, and v the velocity of the film), the record, at the place of such amplitude variation, will consist only of this field having a varying depth of shading, and not so dark as the other portions of the record.

It has hitherto been thought necessary, in the case of a film exposed in the manner described above, to develop it in such a manner that all exposed parts of the film willbe blackened as heavily' as possible. This was attained by employingan extra long developing period, whereby as is well known, the difierences in blackening or shading are greatly reduced. In this way, the

edge zones recorded withrelatively weak inten sity were brought into approximately the same density or depth of shading as the parts exposed with full intensity.

According to my invention, contrary to the previous procedure, I adjust the duration and character of the development of thefilm in such a manner that the film will be blackened less' intensely thanhitherto, and that the intensity of blackening produced will be in direct (linear) ratio tothe intensity of the exposure. With my improved process (employing "normaP development instead of the hard or overdevelopment customary hitherto in sound films) the record will be of the character illustrated by Fig. 3, where each oscillation peak or crest contains a core of greater darkness having'on its edge a fringe the density of which decreases gradually toward the edges of the oscillation-curve. The

same condition exists at each oscillation bottom or trough, where the closely adjacent sides of the trough are of less darkness, and increase in darkness toward the lower edge of the film. If such a film is reproduced, the blending of the shading will cause the surface integrals of trough and crest approximately to compensate each other so that the reproduction of the oscillation will occur according to a curve about as indicated at e. In other words, the gray or partially blackened fringe will cut off the same amount of light as would a completely blackened portion extending out to the line e which closely approximates the true curve representing the sound. Owing to this compensation, the disturbing effect is to a great extent reduced and with it the disturbin subsidiary sounds arising with rapid and sudden fluctuations of. amplitude.

It will' be understood that the division of the slit into a finite number of strips is intended for purposes of illustration only, and that actually the number of such strips is infinite. Therefore the actual latent image and developed film will not have a fringe formed of strips or sections each of which is of the same density throughout; but this fringe will actually be of gradually decreasing density from the edge of the black core to the unexposed white portion.

Briefly, my invention consists in discovering occurs on variable width sound records to such an extent only that the density of each portion of the fringe is proportional to the exposure thereof, so that the total light excluding power of this fringe will be equal to a fully blackened edge strip on the core of a width considerably less than the width of the fringe. In contrast to this, in the prior art, the practice has been to develop this fringe to the same density as the core so that it is in effect a fully blackened strip applied to the edge of the core. In other words, referring to Fig. 3, the edge of the shaded portion represents the effective curve as produced by the prior art, while'the line e shows the effective curve of the present invention.

It is well known to produce sound films in such a manner that, as far as possible, a direct (linear) relationship exists between the quantity of light used for exposure and the quantity of light transmitted during reproduction. However, the desirability of this relation had been recognized hitherto only as applying to sound films made by the intensity process. Sound ,fllms produced according to the amplitude process have hitherto always been produced with non-proportional blacking as the opinion was held that in this method of making sound films there was no intensity effect whatever.

I claim:

1. A process for producinga variable width sound record comprising exposing a soundtrack area to sound wave impressions in the form of an area of uniform exposure and variable width and having a marginal fringe of decreasing exposure merging into an area of substantially no exposure and of complementary form, and developing all of said areas to a blackness proportional to the exposure thereof to preserve the fringe on said sound track.

2. A variable width sound record comprising a soundtrack area having thereon sound wave impressions in the form of an area of uniform exposure and variable width and having a marginal fringe of decreasing exposure merging into an area of substantially-no exposure, all of said areas including said marginal fringe having a blackness corresponding to the several exposures thereof.

ERWIN GERLACH.

Images