Transactions of the Society of Motion Picture Engineers (1916)

of light originates too far from the focal point of the condenser to be of service. It, therefore, becomes evident that there is a limit to the size of light source which can be utilized. There is a very simple method of determining that size. Let us send the light rays backwards, i .<?., suppose the screen were illuminated by another machine or suppose it were made of translucent diffusing glass and lighted from behind, then the ray of light starting at point G on the screen would follow back along the path of the previou ray B. Similarly a ray originating at point H, and so on for every point on the screen. If we replace the arc crater by a piece of ground glass or oiled paper, an image of the screen will be formed on this glass. By moving the ground glass forward and backward, a point will be found where this image will be smallest and its edges sharp. That lighted area then is the maximum size area possible to utilize as a light source. All light rays originating within that area, and falling on the condenser will be projected on to the screen. It is not always convenient to light the screen by another machine and even if it is, the intensity of the reflected light finally reaching the ground glass is low. A more convenient method is to place a piece of translucent diffusing glass (such as heavy density opal) directly in front of the objective lens and hang an ordinary tungsten filament lamp in front of this, (Fig. II). This method of determining the useful size of light source is called back-testing. Let us take a specific example, a machine equipped with a A^y^" diameter, 63^'' focus rear condenser lens and a ^^2" diameter, 7^'' focus front lens and a i^i" diameter, 5X'' focus projection lens. Back-testing it we found the maximum light source size to have an area of .188 sq. in. In practice, this machine operates with 5-8'' diameter carbons burning at an angle. The actual area of the crater is .368 square inches, .18 square inches or 95% greater than the maximum, useful, light source area. This extra amount is, however, considered necessary due to the utter impossibility of keeping the arc crater fixed in one spot, and on account of the shading by the negative carbon (which is most at low currents.) Having this extra area of light source available, slight variations in alignment of the optical system do not materially affect the total amount of resultant screen illumination, though colored borders and ghosts are apt to appear. This, however, is not so with the incandescent lamp. The filament of that light source is of fixed area and remains exactly where it is put. Under these conditions a filament has been designed to fill exactly the maximum useful area ofthe lens system. This area for the lens system at present in use is a square 4-10 of an inch on a side or having a total area of .16 of a square inch. Let us suppose, for example, that in setting up the lamp in the machine an operator fails to center the lamp filament properly in the 4-10 of an inch useful area. Suppose it is i-io of an inch to one side. We then have the condition shown in Fig. Ill, only ^ of our light source is within the useful area. This will result in a very material lowering of the screen intensity. Not only is there a loss of light, due to a portion of the filament being outside of the useful area, but also the beam projected by the condenser is bent to one side, and hence much light is spilled on the aperture plate instead of passing through the aper 49