In addition we had to repeat a large proportion of shots at different camera speeds and with different focal length lenses. From the initial test, it was decided that the 35mm Nikkor would be the main lens to use for most of the shooting. The 28mm was used for certain plates, but the 50mm was generally thought to be a little confining in scope. As a rule we attempted to use the same lens that we would use if we were actually beside a "Speeder" or in the cockpit.
The extremely high light levels presented an exercise in the use of ND filters, especially at the low undercranked speeds that we were using. Toward the latter part of the shooting we were venturing down to four fps and at one point we filmed a couple of plates at our lower limit (at that time) of two fps. Typically we would use a .9 ND on the front and a .3 or .6 gel on the back. At one point we had a .9 glass on the front with a .9 on the back to allow us proper exposure at the two fps camera speed. Generally we tried to keep the same 86ND gel in the camera and only change the glass ND on the lens front in order to avoid the blowing snow on the surface of the gel.
The "average" weather that we were entitled to expect (according to meteorological records going back 50 years or so), was about one out of two days sunshine and clear conditions but that just didn't happen and after the initial two days we found ourselves waiting most of the day, only to "scramble" for less than one mag's shooting before the weather closed in again.
On the glacier itself, the conditions could be extreme and, due to the isolated nature of our shooting, we did not shut down the turbine on the helicopter for fear that it might not start again and we might be stranded when we reloaded or made adjustments.
The closed and heated environment of this helicopter set-up was the only way to function for this type of filming. At one point we had -35C with a 40-knot wind and it took 20 minutes to reload the camera, when normally the operation can be done in three or four. Winds actually exceeded 65 knots on one night, but our hangar tent held. We heated the interior of the chopper in the evening with a small electric heater and the camera system was left on the aircraft at night and heated by an infrared bulb. This provided a very satisfactory solution and could maintain a 30-degree C difference between the interior of the dome and the exterior conditions. During the day the "greenhouse" effect of the dome with its window and black interior usually kept the temperature above +5C, without additional help.
Our initial scheduled period passed and, due to the weather conditions, we were still left with 30 or so plates when we broke for Easter. (It was impossible to film, due to the traditional influx of cross country skiers during this 10-day national holiday.) Until now we had only seen video-taped dailies taken from the screen of a Moviola-type viewer back in London and we weren't by any means relaxing because it really was impossible to get a proper idea of quality from this. During the Easter break, however, I was able to fly to London and view rushes properly.
We resumed shooting after the holiday and, although the conditions had improved marginally, we still had to wait most of the time on the ground. Finally we finished the shot list just a day or so before the agreed cut-off date with the Norwegian authorities and even managed to pick up a few extra static plates missed earlier by the ground camera crews because of bad weather.
The Empire Strikes Back was in every way one of the most interesting projects in which we have been involved. It was especially satisfying to us because it involved the design of special equipment and working in a big beautiful format that increases the visual impact of an aerial camera so much.
In all, we shot really very little film for this type of work in the VistaVision format; 33,000 feet to be exact. The short-comings in the camera motor department have since been rectified, with a digitally controlled closed loop system providing exact camera speeds from three fps to over 30 fps, and other improvements have resulted in what we feel is a very valuable system for aerial and other types of stable background plates, as well as normal production shooting.
