Movie Makers (Jan-Dec 1953)

124 STEREO MOVIES-novelty or nonpareil? A report on the latest 3-D efforts of Hollywood and London . . . How they look and Why JACK E. GIECK, ACL IN his lively article, / Saw Cinerama (Movie Makers, January 1953), John Hefele indicated his enthusiastic reaction to this new entertainment medium. As he pointed out, Hollywood, now scrambling to get aboard the threedimensional bandwagon, is experimenting with a variety of new stereo techniques. Detroit (from where I write) already has been favored with two examples of this new movement; and, after seeing them both, I can report for my part only a lukewarm response. The two samples of stereo fare which currently are touring the country are Arch Obeler's Bwana Devil, in Natural Vision, and a program of five short British films, done in a process owned by Stereo Techniques, Ltd., of England. Of the two the latter is, to my mind, the more satisfactory, for reasons I shall hope to make clear shortly. Both, however, contain serious flaws which seem inherent in the medium, and which I believe threaten the commercial success of stereo movies. The methods of projecting both Natural Vision and the Stereo Techniques pictures are identical. As most of us know, to achieve true three-dimensional imagery (as distinguished from the type of depth illusion created by the large, curved screen of Cinerama) it is necessary to present a separate image to each eye of the viewer. Since all theatres are equipped with two projectors in order to maintain a continuous performance when showing standard "flat" movies, this is accomplished by operating both projectors simultaneously, the left machine projecting the scene for the left eye of the audience, and the right machine showing the scene for the right eye. To separate the two images and "pipe" each into the proper eye, the well known Polaroid process is used. Taped to the glass of the left projection port is an optical filter which polarizes the light from that projector in a plane on a 45 degree angle (upward and to the left) ; a similar filter over the right projection port polarizes the right projector beam in a plane rotated 90 degrees with respect to the left projector (i.e., on a 45 degree angle upward and to the right). The polarized images from the two machines are superimposed on a screen of standard shape and size, but which must have a special metal surface so that it reflects the images to the audience without depolarizing them. Each member of the audience wears a pair of special Polaroid glasses in which the planes, or axes, of polarization are tipped at 45 degree angles, 90 degrees opposed, in exactly the same manner as the filters in the projection booth. Thus, the spectator's left eye will see only the image from the left projector and his right eye only the image from the right machine, since the "wrong" picture, in each case, will be FIG. 1 : Right and left-eye images, carried by polarized light beams of two projectors, are fused in the mind by polarizing spectacles to create single three-dimensional image. FIG. 2: As right and left images cross over each other on screen, the 3-D image seems to move right off the screen and toward the observer. Excess of this effect is displeasing. polarized in a plane crossed at 90 degrees with the filter over the "wrong" eye and will, therefore, be blocked out. (This blocking effect may be compared to the phenomenon most of us have observed when we rotate a pair of ordinary Polaroid sunglasses in front of our eyes while wearing a second pair: we can see through both pairs of glasses until the axes of the lenses are crosses, at which point the glasses held in our hands become opaque.) So much for this stereo projection system per se. However, in order to understand some of the limitations inherent in stereo movies projected by the above system, it will be helpful to examine in more detail how we perceive threedimensional images. The phenomenon of depth perception enjoyed by human beings and a few other mammals blessed with binocular vision (the ability to look at an object with both eyes simultaneously) actually is the result of at least five different kinds of observation on our part: (1) "Seeing around the sides" of nearby objects, by getting a different point of view from each eye. (2) "Triangulating" on an object like a camera rangefinder: actually feeling in the ihuscles of our eyes the amount it is necessary for them to converge in order to intersect on a given object. More convergence is required for near objects than for those farther away; while none at all is needed (since both eyes are directed straight ahead) for objects whose effective distance is infinity. (3) Estimating the relative size of familiar objects: the farther away an object is, the smaller it appears. (4) Noting the relative movement between foreground and background when we shift our position: the background seems to move with us while the foreground passes by. (5) Using plain, old fashioned perspective, or the apparent tendency for all parallel lines to converge at infinity. To this list might be added the modeling effect of light, as well as more subtle factors, such as the fact that more distant objects are seen less distinctly due to dust and moisture in the air. Change in color (bluing) of distant scenery is another help, and undoubtedly there are still other aids of which we are scarcely conscious. Now, it will be noted that only (1) and (2) depend, for their existence, on the observer's having two eyes; (3) and (5) are provided by the monocular eye of any camera, including still cameras, and were used by painters for centuries [Continued on page 131]