Principal Photography

  Q: Did the script indicate precise camera movement throughout the film?

  A: Certainly a lot of camera movement was indicated, but the Steadicam wasn't particularly specified. The camera was just to move, or truck, or dolly here or there. Stanley may have had me in mind from the beginning, however, because the floors were not constructed smooth enough to dolly on at the speeds at which he wanted to move. I was originally supposed to be his wild card for the maze sequences, but I ended up shooting virtually every moving shot in the film because of the floors and because of the supernatural serenity of the moving Steadicam frame. The only conventional dolly shots used were a couple of exteriors parallel to the hotel, when Shelly Duvall ran off to check the Sno-Cat, and one shot that went into the ballroom from the corridor where the wall hid the rails. They laid some rail for those shots, but the rest of it was Steadicam.

  The shot of little Danny Lloyd on the Big Wheel became an inspiration after the fact. Everyone, perhaps including Stanley, failed to realize how impressive the soundtrack would be-and that's what made the shot so exciting. It hadn't occurred to us the wheels would make that loud grinding noise on bare wood and that contrasting silence on the carpets. We discovered it in the dailies because we had put a mike on the camera while we made the shot, but we were too involved in the pursuit of Danny to pay attention to the sound-all these people with their muffled sneakers and muffled curses, trying to keep the damn idiotic wheelchair rig moving fast enough to stay behind him. So we were caught up in avoiding death and destruction and replacing the grips who were knackered with fresh guys lurking along the way who could leap in and start pushing. I was at the apex of this carnival, trying desperately not to roll over in the wheelchair going around corners. We were caught up so completely in what was going on that when we finally settled back and looked at the shot in the dailies we were all stunned. It was great.

  Look at that lens height! Even an Arriflex with a conventional motor sticks down four or five inches. I shot that with a BL, and there was not even half an inch between the camera and the rug. The BL had an 18mm lens, which had to stay essentially level in the set or it would distort all the walls. But if you raise the lens even three inches above the rug, all of a sudden it becomes a completely different shot. The fact that we were below the kid and the vanishing point toward which we were moving was hidden behind him gave this whole sequence a fantastic quality.

  Q: How did you work with little Danny Lloyd as he drove his Big Wheel through the hotel?

  A: It was so easy for him and so hard for us. It was useless to try to slow him down-he just had to go at his own fantastic tireless speed. We just tried to keep up and hoped he would go where he was told. It was like pursuing a miniature locomotive-he was amazing.

  Q: How do you focus the Steadicam on a shot like this?

  A: All by wireless. On that shot, Doug Milsome, who has since become a well-known director of photography (Full Metal Jacket, Breakdown), was with me controlling one of the early wireless focus units.

  Q: Were you in the wheelchair rig for the kitchen tour sequence?

  A: No, I was walking backwards.

  Q: How difficult is it to shoot with the Steadicam while walking backwards?

  A: In a place like that, it isn't very difficult because there a lot of clues to tell you where you are, but in a high-speed shot on flat, open terrain I walk forward and shoot to the rear-the so-called "Don Juan" position.

  Q: Did Kubrick have any preferences about the style of composition?

  A: We had a lot of discussions about the location of the cross hairs. An audience adrift in a big movie screen is not going to know whether or not Shelly Duvall's nose is centered. Kubrick likes centered compositions because they have a certain Palladian quality that he admires. I hated it at first, and then I got to like it. My "eye" had to be reprogrammed to work on that film, and then it had to be reprogrammed afterward to shoot more conventional compositions.

  Q: Stanley Kubrick is known as a director with a high take ratio. How did that impact on your work?

  A: That's where I really learned to control the Steadicam. I got to the point where I could put the frame lines or the cross hairs anywhere I wanted at any time, like a dance. I loved doing fifty or seventy takes, he could not have done too many takes for me because under the circumstances it wasn't very tiring. After each take, there was an equivalent time spent playing it back and frequently a lengthy discussion about the exact position of the cross hairs, so I got a chance to recover. It provided an unparalleled opportunity to bear down and concentrate on technique. Normally, half the work is lugging the gear to the job. through airports in most cases, and the other half is trying to catch up with events and then make a shot based on your own native abilities. If you get three takes, you're lucky, but you never get this kind of opportunity to do a take and see it and do it again and see it again, and to realize that if you keep this shoulder here-this works. If you put that foot there-that works. You cultivate a muscle memory, so you realize that if your hand is just at your belt now, and just at your breastbone now, that you have exactly the boom height you need. I was able to take this wild-haired, wing-it machine and make it into an instrument of real precision, a refined instrument instead of a gadget. A lot of things happen when you have a seminal opportunity like this. Look back at the zoom lens. For all the excesses committed with that tool, the early shooters with zooms had a chance to do something that had never before been done, for good or ill. They had the chance to combine moves and zooms, drifting the focal length in and out. We had similar opportunities to experiment and refine.

  Q: What areas of innovation did you achieve with the Steadicam on The Shining?

  A: We experimented with variations in the technique of tracking shots-particularly in the maze, and during those tours through the kitchen. A camera on a dolly had never been able to take an optimum line around a corner the way a race car driver will. The race car's line is the largest radius you can cut, the closest to a straight line you can get, and therefore comes extremely close to the actual corner. As I backed around through the kitchen, I tried to produce the least disruption to the course of the lens. I needed to maintain a smooth flow of background so the actors would not ever be violently swept across it. The result was a shot with a wonderful serenity, and the compositions blended into one another without sudden rotations. The same thing happened in the maze. When we cut those corners with a 9.8mm lens, the shot took on the quality that those aerial shots used to have in early Cinerama movies when they were flying the viewer up a canyon. They flew as if the plane were a race car and the shot had that same magisterial quality.

  Q: Is there any improvisation in the daytime maze sequence when the Steadicam follows Shelly Duvall and Danny Lloyd?

  A: Yes, quite a lot. The actors didn't know where they were going. We were all always lost in the maze, totally lost. The daytime maze sequence is the great example of those slow-motion race car turns. There was no dislocation, the frame stayed aimed at the actors and the radius of the maze corners had nothing to do with the radius of the camera's motion. The key with the 9.8mm was to keep the camera level, fore and aft. I had to put bubble levels on it to make sure because if I tilted at all, there was hideous keystone distortion.

  Q: You shot the background plates for the speed bike chase through the forest in Return of the Jedi. Why was the Steadicam chosen to do this?

  A: That was an interesting use of the Steadicam. There were three possible choices, and ultimately it came down to an accountant's decision. The traditional choice for shots going through the woods would have been to lay rails through the woods. Only in this case, it would have meant laying thousands of feet in several locations-very expensive and very uneven terrain to work with. Then they would have had to cover the rails with leaves, which would have to be brushed off as they moved along. They would have had to make the shots going forward with a sweeper hiding beneath the lens. They would dolly at max speed along the rails and undercrank the
whole time to end up with the desired screen speed. Of course, the rails would have to be bloody good, smooth enough to allow them to speed up the action, because speeding it up would accentuate any flaw in the track.

  Choice number two was to build a model of the forest and then to push through that on a motion-control camera. Huge.

  Choice number three was suggested by a random phone call from Dennis Muren asking me if I thought I could do the shot. I don't think either of us comprehended how tricky it would be when he made that call, because although you can make a respectable shot going through the forest, you can't speed the shot up thirty times and not expect to see the long-term variations show up looking like bumps. Any slight dip or rise that might take thirty seconds in real time shows up as a wild lurch. So my thought originally was to eliminate those possibilities for error one by one. I would have to give myself long-term stability and not worry so much about the short term. I had already tested an approach in an ecology film I had written that was made by a foundation in Virginia. I drove all around the country for eight weeks, shooting locations from a motor home. I wrote into the film a shot going down various highways all over America taken at two frames a second that I would then cut into a montage. The only narration in the whole film would be over this eight hundred-mile-an-hour rush down roads. When I shot that sequence, I learned that all the normal bumps average out. When you're shooting at one or two frames per second, the little stuff may slightly affect the sharpness of certain frames, but the frequency is way too high for them to appear as bumps. The average point that you aim at on any straight road is always absolutely at the vanishing point, which makes the shot look quite good on average and quite stable.

  Well, I didn't have a straight road as a reference out there in the woods. However, I knew I had to keep the camera aimed at a consistent point. We came up with one technique that relied on a telephoto finder looking at a particular sunlit leaf in the distance. In addition, I followed invisible thread stretched through the woods to keep myself from long-term dips and rises. It got rather arduous, but in the end we delivered the plate that they used for those brilliant mattes. They did a wonderful job on the foreground action, and the whole thing cost a third of what it would have cost for plan A or B, dolly or model. It was just me and Dennis Muren operating a slow-motion, motorized roll-cage on the camera for a banking effect just schlepping through the woods with a Vistavision camera, walking the thousand feet required for each take, stepping over logs, rolling camera at three-quarter fps. Sped up to projection speed, it was pretty astonishing. I was stunned when I saw the completed scene.

  Q: What Steadicam work did you do on The Exorcist II: The Heretic, the sequel to The Exorcist?

  A: I did a shot up into James Earl Jones's mouth, where a tiger jumped out. I did one scene I would never do now and never, ever tell anyone else to do-but I was very athletic in those days. It was a running scene that ended up in a close-up of James Earl Jones's mouth, out of which jumped a matted tiger at the lens. I ran down an alley pursuing a woman through a flapping flock of chickens and wind-born debris. She was to fall in front of me and the camera was supposed to lift up over her. Well, I couldn't quite jump over her, but I ran up a set of steps on the left side of the alley so the camera appeared to rise in the air as I ran up the stairs. Then I changed hands and jumped to a set of stairs on the other side of the alley and ran down them, so the camera just went up and over her. We arranged for her to fall right in the middle, as I jumped from one side of the alley to the other. I never thought a thing about it-amazing shot.

  I also worked on the shots in the African village, and I went to Rio to shoot the film's opening in the slums. In the video version, they restored the sequence-some amazing and, with hindsight, ridiculously dangerous shots.

  Q: How do you work with the sound crew on a film?

  A: Sound guys coexist with us the same way they coexist with everyone else. There have been times when I have actually helped the sound department by carrying a wireless mike because I was the object floating closest to somebody.

  Q: Your vision of what the camera could do if it were unleashed from its bounds led to the development of this remarkable invention. Can you describe what you see when you look through the eye of the Steadicam?

  A: Personally, I like to imagine looking into transparent space, not aware of a frame line. The human eye is not conscious of any enclosure around the space that leads away out of sight. Bumpy motion of the camera produces the illusion of vibration of the frame edge, and this makes a mechanical liar out of the moving "eye." Once the frame edge creeps into your awareness it defines an oscillation, a variation around the edges that is really inhuman to me.

  Good shooting to me is content-driven, and content suggests how the viewer is to be inserted into the scene. It's based on lens placement and a still-to-be-defined art form which considers composition as a dynamic, not a static medium. All of the current rules of composition are written to static shots-I'm interested in the phenomenon in motion. As the camera starts to move, composition becomes fluid because as you move past an object, that has a diminishing weight in the frame. The rules are not of any use in this context because the moment the camera has moved past, the frame becomes unweighted and pops like a soap bubble into a different compositional form. Departed objects continue to have some psychological weight. You sense they are still going by, but now it's off-camera, in space somewhere. The way the frame feels changes instantly as that object becomes unweighted and then is gone. A moving shot may not be centered around the point toward which you are moving, but it may contain that point for a while-perhaps at the frame edge. To me, that's a really interesting shot. Suddenly there's a completely different dynamic to that kind of shot. That shot holds a place static, out of which things flow as you approach it. As you move, everything else, every other pixel in the frame is in radial motion. The mechanical process is not obvious to the audience, but it can have a big impact.

  I also like the ephemeral quality to the Steadicam moves-it can appear to occupy virtually no space itself. Watching ER, for example, you have no idea where the cameraman's physical corpus is. The Steadicam operator is not necessarily at the traditional place behind the eyepiece of the camera, and so the camera moves without the necessity to reserve that clearance-it passes objects like a ghost. It's a thin little wedge that can fight it's way through holes to get a moving shot.

  Q: Is there ever an improper way to use the Steadicam?

  A: Like any other tool for moving the camera, if a move doesn't serve the movie, it's a bad idea. It doesn't matter whether it's done with a dolly, a handheld camera, or a Steadicam.

  Q: What does the future hold for the Steadicam?

  A: Steadicam is a tremendous growth industry at this point. I now believe that it's idiotic to do a movie of any size without a Steadicam operator on hand all the time. It's not a special effect to be brought in, it's as essential as a tripod, particularly on a location film. It is one of an ever-larger group of very clever ways of moving the camera, but it's a key one and one of the best. The art of Steadicam operation by now should be considered a basic survival skill for any camera operator. So I see the Steadicam remaining as one of the basic filmmaking tools.

  In the early days, we always thought some solid-state little black box was going to blow us out of the water. We felt we might have five years before something else came along. I've come to believe now that no black box is going to replace the tripod and no black box is likely to redefine the very fundamental phenomena of Newtonian physics at work in the Steadicam. I think it will be with us for as long as literal film goes through literal cameras. If film continues to be 35mm wide and if motors and magazines are still needed, cameras will only be able to shrink to a certain size and there will not ever be a black box that will move the camera any better than the Steadicam does. If film is replaced by video and everything shrinks down to the sort of camera you hunt for on the floor when it gets lost, like a contact lens-then a
ll bets are off. But if we continue to use real sets and real, nondigital people, then I think this thing has real staying power. We'll grow up to its potential and it will be used very, very well. Excess will come and go. Unmotivated camera movements which are so absolutely, counterproductively, laughable will be in and out of fashion. But the camera will move and must be elegantly pointed. It will move well or it will move stupidly--it's up to us. I think the noble "gizmo" will continue to be an important part of the process for a long time-which gives me joy.

  10

  Fred Elmes

  Fred Elmes, ASC, first became interested in photography when his father encouraged him to use his Leica camera. Elmes took still photographs from grade school through high school. He began to make 16mm movies, borrowing his father's Bell and Howell motion picture camera. Elmes applied to the Rochester Institute of Technology to study still photography and went on to the New York University graduate film school, where he was a teaching assistant. He began to shoot films and learn the craft of cinematography.

  In the seventies, Elmes drove out to Los Angeles and was accepted to the American Film Institute, where he met John Cassavetes and David Lynch. These iconoclastic directors gave Elmes the opportunity to explore significant and divergent cinematic roads. On The Killing ofa Chinese Bookie and Opening Night, directed by John Cassavetes, the camera peers into the soul of the characters capturing performances honed from improvisation and constant revision. The dark and surreal black-and-white images of Eraserhead explore the private world of David Lynch. The seminal shot of the central character, Henry, his piled-up hair backlit and surrounded by glittering particles, defines the distinctive and disturbing tenor of the film.