The Cine Technician (1939)

4 THE CINE-TECH NTT IAN April-May, 1937 Desmond Dickinson talks on BACK PROJECTION (in an Interview) "At Stoll's," said Mr. Dickinson, "we shoot a variety of pictures for a variety of companies. We have had a great deal of back-projection to do, with backgrounds of varying quality, and we have had some interesting problems to solve. A perfect result, as for almost anything, requires infinite care, thought and patience. We have tried to apply these qualities, and our experiences may be useful to others." The main initial problems fell under the following heads : — (1) The "hot spot," or bright patch, in the centre of the screen ; (2) Insufficient light ; (3) Lack of sharpness in the background. These are remedied by introducing a thicker screen, thus diffusing the light more evenly. It might be thought that merely increasing the light on this thicker screen would provide exactly the same "hot spot" as before. But this is not so. Mr. Dickinson uses a 300 amp. arc with a 24-in. parabolic mirror focussing the light down to a circle H-in. in diameter. This intense light would, of course, not only burn the film, but melt the gate of the projector (which is an adapted Bell & Howell shuttlegate camera ensuring absolute steadiness). It is, of course, impossible to have the arc in the booth. It is, therefore, placed outside and the booth given a glass back. The only satisfactory cooling method was found to be an air system passing through a bath of liquid air, and blown on to the film in the gate. The coiled air-pipe in the liquid air bath can be raised and lowered to regulate the degree of coldness. It is thus possible to ensure that the film will not buckle. This is extremely important, as it means a dead sharp picture (in the absence of any other disturbing features). Lack of sharpness is due usually to the back-projection screen being too far away from the next nearest thing in the shot (in a mid-shot this will probably be the artiste). This will have been done because of the necessity of shielding the screen from the spilled light from the lamps illuminating the actors and/or the set. If the projected background can be made intense enough, the spilled light will not have the same de-grading effect. In other words, provided the background print is made with a greater contrast than a normal print, and the light used for projection is very intense, the contrast of the projected background will be similar to that of a normal print with front lighting on the screen. Working in this way enables the cameraman not only to put the artistes near the screen and keep the background sharp but also to light them from whatever angle is correct, even though the spilled light then hits the screen. This, of course, does not mean that you can train H.I. arcs straight at the artistes from the side of the camera ! Having satisfied himself as to the projection, Mr. Dickinson turned his attention to the backgrounds themselves, lie sometimes found those he was asked to shoot were so bad as to be quite unusable. From his experience, he gives the following hints : — Photography of Backgrounds. (1) Use the silenl -film aperture when shooting backgrounds, so as to give the maximum size of frame. This will give, on the back-projection screen, as large a picture with less magnification. This, of course, has bearing, too, on the magnification of grain and the intrinsic sharpness. (2) We have found at Stoll's that stock known as special "background" stock gives better results because it is capable of giving slightly greater contrast with less grain size than ordinary panchromatic stock. (3) In my experience, for stationary backgrounds the only camera that ensures rock-steady pictures seems to be either a Mitchell or a Bell & Howell Shuttlegate. In any case, the camera should be tested by shooting a doubleexposure of crossed lines, the second exposure being taken from a very slightly different angle from the first and the amount of variation (if any) in the distance between any one line in the two exposures checked. This test should be shot on a portion of the roll of stock to be used in shooting the background, since the perforations may not be accurate. To ensure rigidity, it is valuable to insert expanding screwjacks between the tilt and the pan-head (a practice common in U.S.A.). (4) The focus must be pin-sharp. (5) No gauzes should be used. A "beautiful" shot is not necessarily good for back-projection. The same shot, photographed to be less "beautiful" in texture, will be better for back-projection purposes. (6) It is important not to over filter the sky ; a grey sky tends to give a dirty appearance to the composite. For travelling shots (such as from a car or train) which include trees or other objects covering most of the frame both vertically and horizontally, the sky can be absolutely white, in order to give contrast between foreground and background in the background itself. For backgrounds with large expanses of sky, such as wide plains (which do not lend themselves so readily to back-projection), the sky may be corrected slightly if there are woolly clouds breaking it up. (7) The camera should always be level, unless any special angle effect is desired. When the camera is tilted for such effect, the exact angle should be noted with a view to the subsequent placing of artistes and properties in the studio. It is a help, before shooting a background in front of which artistes are to appear in mid-shot length in the composite, to stand people in the foreground in the positions those artistes will occupy, and to shoot a test. This will help greatly in fixing the baseline of the projected background when projected in the studio. (8) It is necessary to get the size of the background correct — e.g., if the background is one of ears travelling, to go behind a foreground car in the studio, it should be known whether the composite is to take in the whole car or only the rear seat, and how much is to be seen around the ear. Processing and Projection of Backgrounds. (9) It is impossible, on the ordinary printing machine, to obtain from any negative a print which is steady to within some lOOOths of an inch. Such machines, of course,