Society of Camera Operators Magazine

Forever Blowing Bubbles
by Robert Steadman, ASC
From the Winter 1993 issue of The Operating Cameraman

     When I look back over the last twenty-five years of my career I would have to say that I am a generalist. That is, I have tried to sample every aspect that cinematography has to offer. Lighting, from table top to massive sets, special effects, overseas work, aerials, high speed cars, marine and underwater are just some areas that I have worked in. However, it is the underwater area that has provided me with some of the most interesting challenges of all. Let me tell you about three film experiences that are all very, very different.

Raise the Titanic

     The noted cinematographer, Matt Leonetti, was shooting Raise the Titanic and looking for a second unit cameraman who could handle a lot of model work, much of it underwater. He knew that diving was an avocation of mine, and also knew that I had a pretty solid lighting background. The picture had been in and out of production for about three years at this point. Stanley Kramer had done over a year of pre-production, including some second unit photography, when he and Lew Grade, the producer, parted ways over some artistic differences. The picture went into hiatus for over a year, and eventually director Jerry Jameson was hired to take over the project. By this time, a massive, multi-million dollar "deep-tank" had been constructed on the island of Malta, and they were on their third special effects crew. The first had constructed the model of the Titanic and some other vessels, including a replica of a 600' Navy salvage ship, a tugboat, and some cut-outs of other ships that could be used in long shots. The second crew had gone to Malta, gotten all of the hardware out of storage and proceeded to get things up and running. However, they could not do much more than shoot some tests, as the deep-tank was not yet operational. Eventually, for reasons unknown to myself, they were let go, and the search was on for another team to make it all work.

     The tank was still giving them problems. It was a huge basin scooped out of the side of a hill next to the Mediterranean, lined with asphalt and fitted with a false bottom to simulate the bottom of the sea. Holding over ten million gallons of seawater, it was a hundred feet in diameter at the bottom and about three hundred feet in diameter at the surface. At its deepest, it was forty feet, a depth where we could work indefinitely without having to worry about decompression. When they first filled it up, a process that took two days, there was concern that it might burst and send all of that water back into the Med, along with whoever might be in it. A daunting prospect, to say the least. It was quickly drained, and plans were made to strengthen the dam that formed half of its structure. Six months later, the work was completed and we were on a plane bound for Malta.

     There is another tank at this facility that has existed for many years, made for the surface photography of models. About four feet deep, and measuring perhaps two hundred by three hundred feet, it was the scene of many a miniature sea battle. At the back of the tank was a spillway that provided a clean horizon line against the sky. We had about a month of work to do here before we were to go into the deep tank.

     John Richardson headed up the new effects crew, and he had a big job ahead of him. The model of the Titanic was constructed on a 1/16 scale, but even at that reduced scale the model was 55' long, and weighed over 8 tons! Heavily constructed of fiberglass, with a deck made of steel, it would not float on its own bottom. Instead, there were massive flotation tanks inside the hull that could be filled with air to adjust the model's buoyancy.

     The first couple of weeks were rather disappointing. The water was cold, really cold. We started in January, and the water temperature was 52 degrees. Standing around waist deep, with the wind machines and ocean breezes on you, was a real test of one's stamina. At every opportunity, we would run to the showers to pour hot water down our suits, but the effect didn't last long, and soon we were right back into it, dozens of times a day. work we would go back to our apartments which were built as summer rentals and had little, if any, heat. After a while the cold started to take its toll. Perhaps half of the divers came down with influenza at one time, and work almost came to a complete halt. The production company, thousands of miles away in Los Angeles, finally realized that something had to be done, and they got us the dry suits that we had been requesting. This made an enormous difference.

     The model was an unwieldy beast. It had been built to accomplish three tasks, and was not really great at any one of them. First, it had to be towed across the shallow tank on the surface. Secondly, it was to be a backdrop for all of the underwater salvage operations, and, finally, it had to break the surface as it was brought up from the depths. Towing was the first problem. It was kind of funny, if it wasn't so tragic, for on its maiden voyage, it slowly rolled over to port and lay on its side, half submerged. The steel deck made it top-heavy, so plans were made to lift it out of the tank for modifications. However, the crane that we had hired proved to be unequal to the task. Just as the keel cleared the water, the crane tipped over and went into the tank, with the model.
Eventually the effects departments worked out their problems and we started to get some good footage. We were shooting this picture in an anamorphic format, using Technovision lenses. All of the surface photography was done at 125 frames per second, fairly standard in this type of work. The cameras were Mitchell Mark II's, housed in watertight boxes with windows fitted for the lenses. The boxes were fitted with flotation pontoons that were adjustable, allowing for the lenses to be within a few inches of the surface. This meant that the viewfinder was also a few inches above the surface, so the waterline came about to your nose.

     Finally we were headed for the deep-tank. The engineering department was satisfied that the dam was not going to burst, and a larger crane was engaged to transfer the model. First, the tank was filled and the model was floated out into the middle. Then it was drained, as the model was guided to her place of rest on the fiberglass and steel sea bottom. Then the tank was filled again.
At the time (1980) there were no underwater housing/camera combinations that would fulfill our needs. We had to be able to shoot slow-motion, but the Arri III was a year or more away. I decided that the Mitchell was just too big, so that left the Arri IIC. The high speed movement would give us 80 frames, and we hoped that this would be enough to give us the illusion of size at 1/16 scale. I wanted to put the package into a tubular housing for eases of construction and balance considerations. A backloading displacement type magazine was found in Germany that would provide the low profile that we needed. From Technovision we got a compact motor base that put the motor parallel and close to the camera body. Since the standard Arri motor was rather a crude affair, requiring a huge rheostat, we decided to buy an industrial motor and build our own electronic control for it. We also built an electronic footage counter and put frame rate and footage on the back of the housing in large LED's. The viewfinder was the next problem. An optical viewing tube would be unworkable for two reasons. First of all, the viewing tube would have to be so long in order to take the image to the back of the housing, it would be very dim. Secondly, it would be impossible to keep your mask up to the back of the housing and do the delicate moves required for miniature work. A video monitor seemed to be the only answer. To my knowledge this had never been tried before, but I couldn't think of any reason that it would not work. We put a 5 inch monitor in its own housing and attached it to the main housing. The system needed no beam splitter because we were dispensing with the optical tube, so the video camera got all of the light. De-anamorphozing was done electronically on the monitor, and we had external controls for contrast and brightness. The widest lens that we could get at the time from Technovision was a 40mm, and it looked through a dome port made from a compass housing. The whole package slid into a tube about 12" in diameter, 30" long.

     Work commenced in the deep-tank. The water still hovered around the 52 degree mark, but now we were fully immersed in it. Since the real Titanic lays at a depth of over 5000', it is in perpetual night. Since we did not have the tank tarped in yet (a massive job in itself), that meant that we were working nights. The model submarines were now operating for the first time in salt water, and they were having their problems. I think that the first night we went down and came back up at least twenty times. It was cold, really cold.

     Exposures proved to be a big problem. We wanted a dark look, but every take got progressively darker until we had not much more than black leader punctuated by the lights on the hulls of the subs. I had shot some tests back in the states, and for these I used a Spectra in its own little underwater housing. This yielded some pretty good results initially, but in the real world we had a host of divers on the set--lighting, grip, camera and effects were all represented. At the end of each take, every diver would, in the course of moving around to do what he had to do, stir up just a little bit of silt. After two or three takes the water would not only be a lot cloudier, but it would act like a filter, reducing the amount of light that would reach the camera. The result was a lot of very dark footage. I agonized over this for some time, and came very close to getting fired until I figured out just what was happening. The solution was simple. I put a spot meter in a housing and made all measurements for the camera position. End of problem. As the silt would start to build up, the meter would sense it and I could adjust the exposure. Actually, that was not quite accurate, for the rather weak lights on the subs meant that I had to shoot wide open - f/2.8 and there was nowhere left to go. If the set got too silty, we would quit and go to another part of this vast tank and do something else.

     The size of the tank caused some other problems. One night, early in the game, I called the surface and asked for a reloaded camera. A diver was sent down with it, but after waiting for five minutes he didn't show up. The staging area for the camera department area was probably two hundred feet from where we were working and I hadn't had any lights installed along the way. He had gotten lost in this ten million gallon lake of ours. A search party was sent out from the set and he was found, thoroughly disoriented by the blackness and lack of landmarks.

     The director of the second unit, Ricou Browning, was a practical joker and so were all of his divers. One of Ricou's favorite tricks was to mess around with your air supply. He would start turning your air off while you were gearing up to dive. Every time you turned away from him, he would give your tank valve a turn. By the time you went in the water it was about a half a turn from being completely closed. Then he would wait until you were engrossed in something, watch for the bubbles coming out of your regulator as you exhaled and give it the final turn. Those that panicked were watched more closely in the future as possible liabilities. Tying your dry suit zipper to someone else's was another favorite trick. When you moved apart the zippers opened and an inrush of freezing water was the result. Another one was tying a bucket to someone's ankle and filling it with air. You went up to the surface upside down, fast. Lots of fun.

     Communications underwater are ordinarily carried out by hand signals. However, we were working in the dark, so this would not work. We had some regulators fitted with ultrasonic communication devices and they did solve the problem for the most part. Ricou and I could talk to each other and the surface and the rest of the crew work hearing-aids so they could hear what was going on.

     Lighting was a whole new ball game. It was not too difficult to make a gloomy, dark look; all I had to do was to make a bank of lights roughly at right angles to the camera, some 20-30 feet back. However, the lights that the salvors used, which in full-size versions towered ten feet high, were only seven inches high in our scale. Somehow we had to come up with small lamps that would light the ship and look like the real thing. Since all of the R&D money had been spent on this long before I came onto the project, we had to come up with something there in our workshops on Malta. MR-16 bulbs proved to be the answer. Originally designed as lamps for slide and movie projectors, they have a small quartz bulb fused in the center of a small reflector. Today they have found wide use in hi-tech track lighting, display lighting and a host of other uses. A mold was made in order to cast the lamp structures out of resin and three of the bulbs were installed in each unit. The wires to power them were hidden in the sand and eventually led to the surface. The big problem was that we had no way to encapsulate the bulbs, which meant that they had to operate directly in the seawater. Their life expectancy was something around two minutes, so we had lots of backups.  At one time we lit all 55' of the wreck and must have used something like 50 of them for one shot.

     The first of Kodak's high speed emulsions, 5294, was not yet on the market, so we were still at an ASA of 100 pushed one stop to ASA 200. This meant that given the brightness of the lights on the submarines, I had to shoot at a 2.8. At that stop, a 40mm lens with a +1 diopter would pull focus at 19", plus or minus one inch. That's right, two inches of depth of field! Actually, this worked out pretty well, since distant objects, such as the ship, were rendered soft by the cloudy water anyway. For some shots I used split diopters and got 28" focus on the other side of the screen.
May finally rolled around and after five months, the water was a balmy 60 degrees. We were finished with this project and ready to go home.