Principal Photography Page 15

  Q: Did you employ multiple cameras on Project X in an effort to capture a full performance from the chimps?

  A: Yes, working with animals you learn getting them to repeat is very rare. Any time you want to do a close-up and a wider shot or a medium shot and get any action to match, you end up shooting multiple cameras to cover it.

  Q: How did you light the vivarium set`?

  A: It was designed with a large skylight. One of the things Jonathan Kaplan and Larry Paull talked about was to create different times of day-early morning, late afternoon, nighttime. The set was restricting because it was so large. It went up to the permanent parts of the stage. It was very difficult to move the lights around up there and to get shafts of sunlight to hit different parts of the wall. So we devised a system with a lot of overhead lights, all aimed at different parts of the walls. They were on switches so we could get the different times of day with light hitting different walls just by turning these switches on and off. The lights were preset. We had a whole series of switches on the wall with a map of the set. For nighttime, we had blue moonlight overhead. We had the practical lights on switches so we could change the mood pretty quickly without doing a huge relight-it was a series of flicking switches. As you get into a complicated film, there are considerations that build in layers of complication. You've got special effects, on top of that you have chimpanzees, and on top of that you have different times of day. All of these are variables you have to factor into any one particular shot or scene. You're always looking for ways to make it as flexible as possible.

  Q: Who Framed Roger Rabbit? is a landmark film in its presentation of live action and animation. What were the challenges in photographing a film where many of the main characters were to be added later during the animation process?

  A: What made it such an interesting challenge was we had been told by Disney there were certain ways to do these animation/live action films. They said, "We've done Mary Poppins, we've done Pete's Dragon. You lock down the frame and let the animator move the character around in it." So they laid out rules they considered to be hard and fast. Of course, there's nothing that intrigues Bob Zemeckis more than an opportunity to push the edge of the envelope. Richard Williams [A Christmas Carol (1971), Rag- gedyAnn & Andy (1977)], the head of animation, was very much in tune and had done all kinds of moving perspective animation. We were going to absolutely arbitrarily violate every rule we'd been given just to see if we could. We shot a test that had a whole list of don'ts that we did. Richard be gan animating to it, and we realized it was possible to move the camera, to have the character move in and out of light, to do dolly and crane moves, as long as the animators knew where the camera was at any particular time. We built floor and wall patterns into the set that were all clues and cues to the animators as to how the perspective was changing, how fast the camera was panning, and all the things they would need to be able to animate effectively to moving cameras. The focal length of the lens you use alters the apparent change of perspective. A wide lens will allow a character to become very small in the background to very large in the foreground faster than if you were to use a long lens, which tends to compress.

  In order to see how animated characters would change size according to perspective, we developed rubber figures to stage the action so that anybody involved with the shot knew where this character was. It was one thing for Bob and myself to stand back and say, "We see the rabbit doing this and this," but it's tough for the actors, effects people, and the camera operator to visualize where the animated characters were at any time-it was almost impossible. The rubber figures were sculpted to full size. We would then rehearse with them. Having Charles Fleischer, the voice of Roger, and other actors performing voices on the set meant we could time the performances. The camera operator could then compose. During the rehearsals, either myself or Bob Zemeckis would manipulate the rabbit or whatever characters there were in the frame. Bob Hoskins would follow and learn his eyelines. The camera operator would learn when to pan and when to tilt on what dialogue he was hearing off-camera. Then we would photograph one of those takes, so the animators would be able to look at where the figure moved, how large he was when he stood back by the door and then when he came up over to the desk. What was the relative size change? We would usually do one reference take. If something went wrong or we got a better idea in that rehearsal, then we would do another take. Then the editor, Arthur Schmidt (Coal Miner's Daughter, Forrest Gump), would be able to understand the intent of the scene, so when they cut together the footage that didn't have animation in it, they would understand where the character was standing at a particular time. Somewhere there exists a strange film of Bob and myself moving these rubber rabbits around and there's enough reference footage that they could cut together a whole version of Who Framed Roger Rabbit? with just the two of us bouncing these rubber rabbits.

  Q: Was there a sculpted rubber figure for every character?

  A: There was one for Roger. We had two or three of the weasels. We had a Baby Herman figure. We had two women who were approximately the right height who stood in for Jessica. We also had a human figure for Judge Doom, Chris Lloyd would actually do the reference for him.

  Q: Did you and Zemeckis approach the film as a film noir?

  A: Yes, we spoke about early detective films, Sam Spade, forties, and film noir. It was an interesting challenge for me. At the time, we didn't know how well they would be able to do the shadowing, tone-matting, and modeling on the characters because up until then, live action and animation films just had flat characters. We accepted flat characters in animation for years. It was amazing when once in a while in Snow White or Alice in Wonderland a character would move in and out of shadow, but at the time, they were always painted very flat. In order to get the characters to look like they were three-dimensional, they were trying to develop a shadowing process. One of my concerns was how effective will that look if we light to very high contrast with the humans and then we put an animated character in? The shadowing and contrast had to match. The lighting process became a fine line between a film noir look, a high-contrast and stylized lighting, and a not so contrasty look so the animated characters wouldn't stand out as not being part of a three-dimensional world. When we started working, we had not seen the tone-matting process. It wasn't until towards the end when we started to get animated material back that they were developing a process. Then we began to see how much they could do shadows and highlights on the animated characters. So it was a real tricky process to pre-visualize how much they could move in and out of shadow and how much shadow and contrast you could build into a shot and still not be too much for the animation process that was coming later.

  Now, the techniques developed for Who Framed Roger Rabbit? carry through as you watch animated features like Beauty and the Beast and The Lion King. Animators don't animate flat anymore, there are tremendous perspective changes. Now, they computer generate backgrounds so the apparent camera moves through the animated landscape and the characters all have highlights, shadows, tone-matting, and all of the things Who Framed Roger Rabbit? started. It was pushing the envelope, creating something the audience now accepts as part of animation language and grammar. Now, if it isn't there, then it becomes something that's missing and looks false to an audience.

  Q: The ship sequences in Hook were shot in the studio. How were you able to create daylight on the stage using artificial light?

  A: Steven Spielberg and I had initial conversations. They were going to build a ship and the town on an island in the Caribbean. Then Steven realized it might look too real. What he wanted was something between a theat rical fantasy and a real look. The decision was to build everything on stages which would give us more control, but also the challenge of creating the illusion they were outside.

  Steven and I screened Greystoke: The Legend of Tarzan, Lord of the Apes, which had effective jungle sequences. They had shot some of it exterior and some of it on the stage. We analyzed what made it look real
. In the early days of Hollywood, they often would build a set on a stage that was supposed to be an exterior. They had control over all the lighting and could make everything perfect. They made exposures of the contrast ratios from the shade and the shadows so the sunlight fit the film, especially in color films. The early Technicolor days were dictated by contrast ratios. So audiences got used to seeing two kinds of exteriors. The ones on the stage looked phoney, but in reality everything was very controlled. When you go out in the desert, it's very bright sunlight and very dark shadows. So you have to compromise. A lot of times you expose more for the shadow areas, and the sunlight and the highlights get very hot. Sunlight is brighter than film can accept, so the audience is seeing burned out highlights and overexposure in certain areas, and the shadows are darker. Steven and I decided the sunlight coming through would be overexposed, hotter than what the film would want to see, and also tried to build in a single source. Often, to get as much light, you have a lot of lights on the stage and you get multiple shadows on the ground which the audience interprets as being incorrect. The trick was to find ways to create the illusion it was just the sunlight creating the light. The rest of it comes from the normal process, which is the ambient skylight, by using a very soft fill light. So we built a very hot sunlight into lighting of the sets and on the deck of the ship. One of the difficulties was, anytime we tried to get a wide shot, we always ended up seeing the limitations of the stage; where the sky ended and where there was no ocean past the ship. We accepted that as part of the theatrical, fantasy feeling. They were in a land that didn't have all of the same rules we do.

  Q: How do you and Steven Spielberg work with the camera?

  A: Spielberg is a visual storyteller more so than directors who deal strictly with performance. So his eye is really trained for that kind of visual evaluation of the frame and composition in addition to performance. He has said he feels the film is being made within the frame lines in the camera. Over a period of time, I've worked more and more with a remote head. It's a process of putting the camera on a motor servo-controlled head. Then, instead of physically looking through the camera on the dolly and operating the head there, you are looking at a video image and operating the head remotely at a video monitor. A long cable runs down to the end of the crane arm or wherever the camera is. The image is larger. Instead of looking at a very small image through the camera, you are actually looking at a pretty good size image in front of you. There's just something about the effectiveness of being able to watch with two eyes on a monitor as opposed to one eye, squinting. You can evaluate, your eye can click around the frame and watch what's happening. You can actually do much more complicated moves than if there was a guy looking through the camera. You can do 360 degree pans and complicated dolly moves. Spielberg used remote heads to a certain extent on the Indiana Jones pictures. Bob Zemeckis uses it constantly. I would say half of the film is shot with a remote head when I work with Bob Zemeckis. Steven tends to design complicated shots like low angles that sweep along the floor, then up over the table and across the room-shots that are most effectively done with a remote head. On the past several films Steven and I have worked on, the majority is done with these heads. Spielberg's reflexes are good. He's operated for many years, and has developed those reflexes. Steven is one of the few directors I have worked with who operates. Ray Stella still operates most of the time. In the case of Bob Zemeckis or with Ron Howard on Apollo 13, the operator operates the camera because they have all these reflexes they've developed over the years, but Steven is actually able to adapt very quickly to operating. We recently gave Steven an honorary operator's card because he has so frequently operated and has so much respect for the operator. Steven feels the film is made by the operator. A lot of directors don't feel comfortable operating because it really takes a very trained set of reflexes to look for little visual cues of reacting in advance of when an actor is going to move. There are all kinds of reflexes an operator has, and an instinct of when to stop the camera because the shot is properly composed. How much headroom do you give an actor and where do you place them in the frame which balances visually without putting them right in the center? Most of the time, the operator is ignoring the dialogue or the performance because they have other things to evaluate. Most directors prefer to evaluate the performance rather than being worried about panning too far. Wisely, most directors don't fool with that. I know if I operate the camera, I suddenly have to shift into operator mode and watch for the sandbags, lights, and composition. Sometimes I don't get to watch if the actor landed in his light properly. My job as director of photography is somewhere between evaluating as a director and as an operator.

  Q: How did you photograph the sequence in Jurassic Park where the two children, played by Joseph Mazzello and Ariana Richards, are trapped in a kitchen by the dinosaurs?

  A: The kitchen scene is the most effective use of storytelling and the dinosaurs in the whole movie. Instead of the dinosaurs just being in the background, they are an integral part of the action and the drama. Steven Spielberg was very concerned with controlling the cost and schedule on Jurassic Park, especially since we were dealing with a lot of unknowns in the effects work. They storyboarded the film, principally the action sequences, but not just in the conventional way of a storyboard artist drawing sketches. They made a computer-generated video storyboard. They went to the trouble to construct a T-Rex and the Rapiers, and they were able to move the camera and actually do the shots in a very crude, primitive way to see how effectively they would work. They also intercut static shots of the drawn storyboards so we would end up with an animatic of whole sequences. We could look at the timing and at how effectively the scene worked. When we actually shot the sequences, we would run the animatic to see how the script had dictated and interpreted the action. Steven had a great deal of input on the animatic, and it became the very serious guidelines we followed. There were times where we said, "This worked pretty well, but a really good angle would be down here," because now we had the real set, we could see a really interesting, dynamic angle. So we would modify as we felt was necessary, but there were particular shots that worked real well in the animatic. If there had been two or three iterations, they would find a very effective one and we would as much as possible do those shots. There were some really effective moving shots for that kitchen sequence. One of them was a very low angle on Joseph Mazzello, and the camera dollies around as the Rapiers come around behind him. Then it becomes a real low angle, looking up, and the Rapter lifts his head up and hears Ariana at the other end. We had been doing other shots typically with the heads of Stan Winston's (Edward Scissorhands, Terminator 2: Judgment Day) creature puppets, but this particular shot required a movement, the head turn, and stepping forward that could only be done by the computer. So that was a case of analyzing the move which would allow the space to put in the computer-generated creature, and understanding the timing of when you would hear Ariana. Then the creature would raise his head and we would have to tilt up with him. Ray Stella had so effectively learned to frame for things that weren't there with Who Framed Roger Rabbit?, he was able to visualize in the frame when he had to move and how high he had to move. We would rehearse, again taking the Who Framed Roger Rabbit? reference technique. We had full-sized rubber Rapters and other creatures we would rehearse with, so everybody knew how much room to allow for a creature that size at that distance. Again, we would frequently shoot one take with these rubber Rapters with someone just walking along behind them in order to have the reference material for the animators, even though they were going to be using a computer to do it with reflective changes.

  There are times when you have to do a moving camera shot with a motion control camera and a computer because the information rate of panning and tilting the camera and of the dolly move will then get recorded by the computer for it to generate the movement of the creatures. With the computer, we're less and less restricted by what we had to do in the past by putting in frame reference po
ints like little fluorescent dots on the floor, the set, the props, and on furniture pieces and using a motion control camera. Now, those have all become reference points for the computer. We measure from the camera to the dots, then we'll make a chart of which ones they are and the angles. So that becomes information between the animators input into the computer for reference of the camera moves. On that shot in the kitchen, it was a case of rehearsing and visualizing what was there, then just shooting empty space and air. The animators very effectively put in the dinosaurs later.

  Q: You have worked on so many complex special effects driven films. How have you learned about the technology involved in special effects photography?

  A: One of the things that's so fascinating is there are no courses for what we do. We can give them after the fact, but at the time when somebody goes up to make a film, everybody's just out there learning how to do it. When we started Jurassic Park, the intention was to use all stop-motion dinosaurs and conventional effects. So at that point, yes, any of us could have taken a course in stop-motion dinosaurs from anybody who had done King Kong (1933), all the way up to contemporary times. It came down to somebody who said, "I wonder if we could do this with computers?" They did a couple of tests and showed it to Steven. Just before we started, Steven said, "Do you think you really can do this?" and everybody said, "Yes, we can," not knowing whether they could or not. So Steven had the courage and confidence, having worked in the past with these guys who had said they could do something and then delivered. He also fortunately had the clout to be able to go to the studio and say, "We're going to dedicate millions of dollars to do something that nobody has ever done before and nobody knows if they can do it," and the studio said, "Okay" So, we started the film not knowing if any of the special effects could be done, and throughout the film they developed the technique.