Within a few days the 4-perf and the 8-perf crew had melded into a basic camera unit. Seven and eight cameras—three of them VistaVision - were shooting at once. I loved it! It gave new meaning to the old multiple camera dilemma of matching light. But if the cuts are exciting enough, it just doesn't matter.
Between Richard Edlund, Alan Hume and myself it was a frenzy of shooting, very exciting z but long dailies. All the operators had access to the storyboards for each major cut and at least one camera would shoot that angle. Many times we had a VistaVision camera shooting our master and the 4-perf cameras sprinkled here and there for coverage. That was one of the remarkable things about the production: it had strict guidelines and storyboards, but as long as we did the required boarded shot we were free to shoot any other cut that might possibly be usable. Richard Marquand never tired of checking potential angles and in this territory of unboarded material George Lucas was there, always available with his vision. The pace was fast and began to become suspiciously like fun.
It was always a treat placing those other cameras. I try not to think in terms of camera angles but rather in placing the audience in an interesting position, preferably in a place they've never been.
So often we take a set or location for granted after we've been on it for a few days and the audience only sees what we show them. I know this sounds like convoluted semantics but using the perspective of the audience as a guide to camera placement produced very good cuts.
Some of the most exciting cuts were hand-help: it put the camera right in there with the action. For strong action material the anamorphic format is wonderful with the hand-held camera—the wide screen is more powerful on the frame sides with foreground movement—and more intimate.
In and around the Sarlacc pit much of the coverage was hand-held with the operators on ropes or wedged in behind a rock. It was the fastest way to get the best position. Of course it was hard on the operators and, with all that blowing sand, brutal on the cameras and the assistants.
It is a tribute to Arriflex and our assistant cameramen that we had only one camera jam and one piece of sand-scratched film during the two weeks we were in the desert. It was Michael Benson's turn in the pit for a hand-held shot—dangling on a rope down by the Sarlacc's teeth. The wind increased and caused a minor sand storm in the pit—a swirling vortex of powdered sand covered Benson completely. Afterwards a plaintive cry was heard. "Send down another camera." In eight hours the jammed camera, an Arri IIC, was completely stripped down to the frame by Michael Glennon, regreased, reassembled, a steady test shot made, and when the lab reported "O.K." it was put back on the line.
Tom Laughridge was intrigued with finding an angle showing the barge and one skiff without showing the skiff's supports. After about an hour he gleefully reported that by putting a camera in one of the lower side ravines and using a foreground sand dune to mask the supports of the skiff, the illusion of a floating skiff was quite good. Skeptical at first, George Lucas checked it out—generally shots like this were reserved for VistaVision where the supports could be painted out. The floating illusion was so perfect he accused Laughridge of having ambitions for the visual effects department.
Richard Marquand was most comfortable working directly with the camera operators. Commonly known as the British system, it places more responsibility on the camera operator and requires a great deal of input. In the credits the camera operators are referred to as operating cameramen, a title reflecting their importance and responsibility on the shoot.
The expected rain at the Crescent City redwood locations never appeared. We were cursed with bright sun, except for morning fog. This was the acid test for my theory of shooting ridgetops and high slopes early in the morning and low forest areas at mid-day and afternoon. The theory worked fine and for the occasional sunny spot we'd set up a 20X20 silk or 12X12 grifflon to block the sun. The camera operators had to constantly avoid bright sun beams on forest plants. Our fast shooting speed is a credit to the planning of the production department and their scheduling the right locations for the right time of day.
Various forest locations were spread out over a five mile area. To make good use of the early morning light, sometimes the ILM unit would split off and shoot a separate board. I would send a 4-perf camera with them to pick up added cuts of action primarily staged for a VistaVision camera. Sometimes we sent Tom Laughridge out with a separate unit to pick up shots of action in a forest locale inaccessible for the whole company.
Again, if one of the ILM operators or assistants were free he would come over and start shooting with us. On several occasions during the forest battle sequence we had more usable cameras than operators. I would not hesitate to let an assistant operate. I think every assistant on the show had a chance to do at least one shot at operating, including the British camera assistant, Mike Frift, who has a very good eye for action.
After two weeks another unit was started at ILM in San Rafael, which required the principals for blue screen process work. Alan Hume, Richard Marquand and George Lucas left to supervise that unit, while David Tomblin, first assistant director, took over the direction of our unit. We were shooting the battle sequence and preparing for a grand shoot-out between the Ewoks and Imperial Storm Troopers. It meant three weeks with hundreds of actors and some pretty snappy special effects.
[ continued on page 4 ]