Journal of the Society of Motion Picture Engineers (1930-1949)

marked elongation, an advantage when designing spectacular scenes. The backgrounds of normal interior sets are not likely, however, to suffer any appreciable deformation. It is the foregrounds, especially when they contain objects of familiar shape, which may be more visibly distorted. An example of a visible kind of stereo distortion is worked out from the data on the following shot. Practice, and frequent viewing of 3-D films, will tell the stereotechnician what is acceptable and what is not. It is un. likely, however, that he will satisfy everyone; for reasons that are not yet clear, people differ enormously in their sensitivity to stereoscopic shape and size. When the stereo settings have been made, one more step is required to be taken before the camera is ready to roll. This is the Stereo test, which provides the necessary data under the microscope to determine the actual, as contrasted with the nominal, values of cot <p (or h) and tc. To make this test, a small target board resembling a ping-pong bat is run out first to a distance of 40 p and then to 22 p with the aid of the Stereotape, a few frames of film being exposed at each distance. The lens focus is set at 15 ft for both shots, since the lens-to-film distance enters into the equations, and must therefore remain constant. The shot which follows Slate 35 in the film had already been taken. It was the nearness of its front banner which had to be exceeded in 35, in order to produce the desired progression in space. The entry in the stereocontinuity book is as follows : Slate 15: Shot opens with pair of banners in CS, filling screen, at 79p (6 ft 4 in.). Banners raised out of picture to reveal further pair, and so on, till raising of 4th pair reveals Odile (Beryl Grey) in LS, who starts to dance her Variation. Camera static, rear of set at 13p (39 ft). At end of slate, Odile has danced into MS at 48p (10 ft 5 in.). Treat shot as 5 + , with/c = 50 mm, tc = 1.25 in. and cot <f> = 220. Thus Do = 13p, D! = 44p, and D2 = 75p, since AI = 31 p. Hence front banner is slightly closer than N2, actually #2.13. The method of working out this shot need not be repeated here, since it resembles the previous example and can be checked with the help of the equations already given. Although the stereotechnician will seldom have to force the cameraman's hand in the choice of lenses, or the director's in the arrangement of a scene, he is nevertheless bound to be constantly preoccupied with the smallest value of tc which his equipment will provide. In studio work, with its large depth ranges, he is likely to be pressing against this limit much of the time. In Slate 15, for example, 1.25 in. was the absolute minimum tc available with the 50-mm lens, and had it been desired to hold the nearest banner farther away than NZ.I, while retaining it at the same field size, nothing could have been done — except by allowing divergence to occur in the farthest planes of the shot. Image Distortion in the Theater The data already provided for Slate 1 5 makes another interesting analysis possible ; by studying the shape of different parts of the image, it is possible to get a clearer idea of the distortions set up in the motion picture theater. Let us consider the plane in the image corresponding to Beryl Grey's position when she has danced forward at the end of the shot, and is at 48 p. The spectator is assumed to be at a distance from the screen of 2.5W. Example: It is required to find the depth magnification, width magnification and shape ratio for Slate 15 at a plane in the scene distant 48p from the camera, when the spectator is seated at 2. 5 W from the screen for which the film was shot. Spottiswoode, Spottiswoode and Smith: 3-D Photography 269