SHOWSCAN - DISCOVERY II
Camera Assistant, Susan Walsh, left, preparing to Pilot Deep Rover on a tethered dive. Right, examining some of the local sea life.
Being asked to photograph a film in Canada, for Expo 86 is an interesting assignment. Add to that the idea that it is an underwater film and it becomes even more interesting. But when the offer comes from special effects genius/filmmaker Douglas Trumbull, who tells you that you will be shooting this film in his revolutionary SHOWSCAN (tm) 70 mm, 60 frame-per-second process, what you have is a dream project.
Thus began a six month effort which was one of the most interesting, technically challenging and physically demanding projects that I have ever undertaken. The results will be shown for the rest of this year at Expo 86, in Vancouver, B.C., in the form of a simulator ride, or "ridefilm", called DISCOVERY II.
Trumbull's Showscan Film Corporation had a contract for three special venue films for Expo 86. Two of those films were to be produced by Canadian companies with Showscan providing technical support, but the third was to be co-produced by Showscan and the Canadians. It was for that film that I was hired as the Director/Director of Photography. It was to be a point-of-view ride, a simulator, like the Star Wars ride at Disneyland. However, this ride would be in a deep submersible submarine during an emergency, undersea rescue operation. The purpose of the simulation was to promote Canadian expertise in undersea technology. They wanted to highlight CAN-DIVE SERVICES, LTD, a large Canadian commercial diving operation. In particular they wanted to highlight Can-Dive's sophisticated, one-man DSRV, DEEP ROVER.
DSRVs represent the Buck Rogers future of submarine technology, but they are here and working today. Spurred on by the oil industry's demand for men and equipment that could do useful repair and maintenance work in deep open water, the designers of these machines have found the financial incentives to conceive and build the most fantastic devices, utilizing space-age materials and techniques. These submarine vehicles are just coming into their own in recent years. With such tools, we are finally beginning to really be able to work and explore in the abyssal depths that comprise so much of our planet. Witness the recent, fantastic discovery of the wreck of the Titanic which, for lack of technology, has so long eluded us. The finding of that ship, and the photographs taken, prove that no depths are too great for man to penetrate and explore.
For Expo 86, Showscan would design and build the simulator and its automated sound, projection and video monitoring systems. The simulator interior would closely resemble the actual interior of Deep Rover. Showscan would also produce the POV film footage comprising the exterior "reality" being seen by 25 to 30 "passengers" through the large spherical view-port, a central feature of the simulator. They would also produce video footage, sound track and graphics for the simulator's multiple radios, instruments and monitors. The simulation would start when a tarpaulin is removed from the view-port, just as we are being lifted off a ship's deck and placed in the water. It ends when we're lifted back on deck and the tarpaulin is placed back over the view-port. Total time for the "ride": 6 minutes. Total screen time for the 65mm Showscan exterior "reality" portion: 4 minutes and 45 seconds. I thought it would be relatively easy and straightforward. I was wrong!
It was May of '85 when I began the first scouting trip. The first phase of the trip was to meet with all of the key people who would be involved in the project on the Canadian side. We needed to take a look at the subs and other equipment that we would be working with. We also needed to discuss our mutual problems and needs with respect to the filming of such a complex situation. The second phase was to begin actually diving the various potential filming sites to assess their relative merits. In our first meeting, May 24th, in Vancouver, it quickly became clear just how logistically complex our undertaking would be.
Leonard Gmuer, SHOWSCAN'S UPM, and I left LAX on a 7:45 am flight and were pulling in to CAN-DIVE'S maintenance yard located in the North Vancouver waterfront area at noon time. In addition to submarines, CAN-DIVE was to provide the technical and logistic support that would be required.
While at CAN-DIVE we met Alan Scott and Rick Hinton, from the EXPO '86 Pavilion team, Jim English, the project manager for CAN-DIVE, and Dave Griffith, an underwater Archaeologist and president of the Underwater Archaeological Society.
The first thing I did was look at Deep Rover and Sea Otter. Deep Rover looks like some kind of space capsule. It is basically a plastic sphere sitting on a base that contains power, propulsion and life support. It also has highly articulated robot arms, which make it capable of practical work at great depths. It is certified for 3,500 ft. with a crush depth of 12,000 ft. while its pilot sits in shirtsleeve comfort at sea-level pressure. It will cruise at 2 knots with a top speed of 3.2 knots. The life support duration is 150 hours.
Sea Otter is an older, larger and heavier 'YELLOW SUBMARINE' type of submersible, with a sausage shaped steel hull and small portholes around its nose and the small conning tower. While it is only certified for 1.500, feet it will carry a crew of two or three and with its single robot arm; it too can do useful work at depths far too great for human divers.
Jim English gave me the cook's tour of the subs and within a short time we understood each other's needs very well. We met for a couple of hours, went over the script and storyboards to familiarize everyone with our needs, and began to make estimates of what equipment and personnel would be needed. CAN-DIVE is a very professional operation. They do things like under-sea construction, maintenance, salvage, inspections, pipeline work and so on. They have a staff of commercial divers, tenders and technical experts in all the crafts necessary to support a major diving operation. By the end of our initial meetings, I was grateful that CAN-DIVE was our partner in this undertaking. I knew they had the expertise we needed to make this project happen.
When all was said and done, we would need four submarines: Deep Rover, Sea Otter, Hysub and Manta. To handle and tend the subs, all of which weighed from one to five tons, we would need the 100ft salvage ship, Redonda, with its two large, high capacity cargo booms, and Manta's 40' support vessel, Sea-I. For a camera support vessel, we would need the 60' Sea Tech. When the camera was in its normal 400 lb. configuration, Sea Tech's small cargo boom could handle it. However, when the camera was in its 1.5 ton "Big Dome" configuration, it would be handled by one of Redonda's booms.
For crew quarters we would need three 60-foot boats: Blue Fjord, Clavella, and Oceaner. Since we would be working miles from the nearest port, we would need two fast 30' cabin cruisers, Stel Star and Rye Whiskey. They would ferry equipment, supplies and personnel to and from the location. The total number of people needed on the shooting location, including all of the divers, sub drivers, boat captains and crew, technical support and of course the camera crew exceeded 40. We needed underwater welding equipment; surface supplied diving gear, a 400-amp generator for underwater lighting, and a 400-amp generator for underwater cutting and welding equipment. We also needed several ATCO trailers with heat and light for use as workshops and camera department on the ship's deck. In addition to all the miscellaneous diving equipment and supplies that would be needed and provided by the Canadian team, I would also bring 45 cases of camera and diving equipment.
SCOUTING FOR A LOCATION
As far as locations were concerned, that was Dave Griffith’s area of expertise. That's what he and I would be scouting for in the next few days. Our requirements were that we wanted a bottom area, preferably no deeper than 40 to 50 feet maximum. It had to look like a rocky shelf at some great depth in the lightless abyss. However, if it was much beyond 33 feet (the no de-compression limit for air divers) we would start to encounter serious limitations on bottom time and thus on shooting time. I knew we would be spending a lot of time underwater and couldn't afford to deal with diver de-compression along with everything else. We also needed a shipwreck and interesting marine life, which were to be seen as we descended for the rescue operation. Then there were weather and water currents, visibility and other environmental factors to be considered, as well as production deadlines. According to Dave and others, we would have our best success in an area 250 miles north of Vancouver and thirty miles north of Port Hardy, the northernmost village on the tip of Vancouver Island. It is the gateway to Queen Charlotte Sound.
British Columbia waters are some of the most interesting diving in the world. Nevertheless, for most of the year weather considerations make a project of this type difficult. According to our consultants, the best time for an all-round combination of good weather and visibility was the end of August to mid-September. However, a window of opportunity was statistically likely to open in July, giving better water clarity and good weather at the same time. After September, visibility is the best of the year, but weather is very unstable and can deteriorate rapidly. We needed the best possible visibility, but could not risk the near certainty of winter weather. We decided to gamble on the possible July window.
By 4:30 that afternoon, Leonard, Dave and I were lifting above the waters of Vancouver's Coal Harbor in the chartered Cessna 185 Floatplane. We were heading for Port Hardy, in the Queen Charlotte Strait of northern Vancouver Island.
It was a wonderful flight. The terrain in this part of the world is pure North Woods. Wherever there is land, there is forest as far as you can see among the countless islands that make up the coast of British Columbia. Cruising at 3,000 feet, we headed northwest up the Strait of Georgia. To our left was the vast mountainous bulk of Vancouver Island. Beneath us was the endless labyrinth of islands, straits and inlets with names like Discovery Passage, Surge Narrows and Alert Bay that are a vital shipping link between Vancouver and points north. Altogether it's about a 250 mile trip, and for about 100 of those miles the straits between Campbell River and Alert Bay are narrow, anywhere from a few hundred yards to a couple of miles wide. When the tide shifts among these passages, it creates currents that you can quite literally waterski on while anchored to some stationary object on shore. In many places, you can see whitewater that would make any good river rat jump for joy. However, it is no joy for the sailors who must venture through these same waterways for their livelihood and it was currents like these that we too had to avoid. We continued our flight northward for about two hours and arrived in Port Hardy with a lowering ceiling and about a mile of visibility. Even so, our pilot was able to sense when to turn off Queen Charlotte Strait to line up on final for the runway at Port Hardy.
Port Hardy is, as its name implies, a hardy logging and fishing village. It's the end of the line for the two lane blacktop, which runs the length of Vancouver Island. It is also the hopping off point for sportsmen, vacationers and a lot of people who make a living from this wilderness of ocean and forest.
We called the local cab for a ride from the airport to the main boat docks in town, about two miles away. There amid the aging fishing fleet we found the 60' Clavella that was to be our scouting vessel for the next three days. Owners John and Joann Deboeck greeted us on the dock and helped us quickly stow our gear. Then, as night fell, the five of us walked up the long dock toward town and dinner.
The next morning I awoke to the sounds of Clavella's diesel engine chugging us out of the slip and into the waters of Goletas Channel. It was low ceilings and low visibility in light rain and fog as we headed for an inlet between Nigel and Balaklava islands, about two hours away. Dave Griffith knew this particular location held promise because of its proximity to an area in the channel between the two islands known as The Wall. There was also a shipwreck in the area, the 280-foot steamer Themis. Finally, he knew of a small cove just adjacent to the channel, protected from most currents, which he thought might serve as our main set.
Dave was right! The area was as close to perfect as it could be for our needs. The EXPO people wanted the film to show some of the interesting and diverse marine flora and fauna and The Wall certainly did that. As the name implies, it is a vertical rock wall, which borders one side of the channel between the two islands. It starts as the surface and plunges directly to 80 or 100 feet before tapering to a gentle slope near the bottom. Since it is in a narrow channel the constantly shifting tides keep the water moving at two to five knots. Because of this constant movement of fresh, nutrient rich seawater through this area, the abundance of sea life growing on The Wall is extremely unusual and stunningly beautiful. It looks very much like a densely woven tapestry designed by King Neptune himself.
We spent the next two days diving on The Wall; the wreck of the Themis at the mouth of the channel and the cove, which would become our set, where Sea Otter would be trapped. By the time we were done, I had measured depth, taken light readings and still photos of much of the area. The area we would use for our set was particularly well suited to our needs. Since the subs needed to be tended by the 100' Redonda, we needed to be able to moor the ship almost directly over our set. The bottom geography of the cove would allow us to do this. According to the storyboard, the Sea Otter is supposed to be at some unspecified but very great depth, in a rocky area with gullies and cracks in which it becomes trapped by the heavy cable it is inspecting. Our cove, called Port Alexander on the charts, had a rock wall that descended almost straight down to a rocky, sandy bottom at 35 feet, and it was well shielded from the constantly shifting currents.
THE NEED FOR DRY SUITS
After three days of diving/scouting in the 40 degree waters and gray, wet weather, another important consideration became apparent: the need for greater warmth than my wet suit could provide. I was diving in my heavy-duty (by California standards) Lycra suit, which consists of 1/4" farmer john pants and 1/4" one piece jacket with hood attached. Both garments have nylon on both sides and when fully suited you're protected by a half inch of rubber and four layers of nylon from your crotch to your neck, and a quarter inch of rubber and two layers of nylon on arms, legs, hands, feet and head. I have dived many times in 40-degree water with this suit and felt adequately protected. The operating principle of the wet suit is that the snug fitting suit traps water very close to the skin and your natural body warmth will quickly warm the water, thus keeping you quite comfortable. In addition, the rubber itself is a good insulator.
However, for extensive diving in waters as cold as this, a dry suit is a necessity. Unlike the wet suit, the dry suit (as its name implies) keeps you dry and therefore warmer. It is made of non-porous rubber which seals tightly at the neck, wrists and ankles. Unlike the tight fitting wet suit, which is a struggle to put on and take off especially when it is wet, the dry suit is very loose fitting. It is easy to put on and take off through a single, large, watertight zipper. Underneath the wet suit, you wear bathing trunks or a body stocking. Under the dry suit, you can wear anything you want, depending upon how cold it is and how warm you want to be. But basically you wear the kind of clothes you would wear in any cold weather: thermal underwear, sweat clothes, bulky wool sweaters and socks, Levis, whatever makes you comfortable.
On the scouting trip, we were making many dives each day in cold, wet weather. Naturally, between dives we wanted to shed the diving gear and go into the warm cabin. My Canadian friends had dry suits and this allowed them to dive with greater warmth and then quickly and effortlessly shed the garment, leaving them dressed in warm sweaters and pants. When the time came for the next dive, they simply pulled the dry suits on over their clothes and they were ready. For me, however, it wasn't so simple. My wet suit performed admirably while I was in the water during the relatively short duration scouting dives. However, I knew that it would not be adequate for the long periods of submersion and relative inactivity we would experience during filming. During the scout, the worst part was putting the suit on and taking it off on the cold, wet, windy deck. It was sheer hell standing shivering in my Speedo trunks, turning blue and growing goose bumps the size of goose eggs, while the others were already below, drinking hot soup. After the first dive on the first day, I would have given anything for a dry suit and I vowed that my underwater crew and I would have the best dry suits available when we came back.
I returned to Los Angeles armed with a pile of location stills, tide and current tables and charts of the area, as well as lists of things we would need when we began shooting, including dry suits!
Susan Walsh, with whom I have worked on many projects both above and below the water, was to be my key first assistant and right arm for the duration of the project. Doug Trumbull's technical assistant, Barnaby Jackson, who lived and breathed 65mm cameras and Showscan cameras in particular, would be my camera technician/engineer. We had about seven weeks to prep the job. Normally that would seem to be an excess amount of time, but in this case it was barely enough because we still needed to re-design the existing Showscan underwater housing to meet the very special needs of this simulator project. In addition, we had to find and purchasing appropriate dry suits and other special equipment that we would need.
The dry suit problem was easy to solve. After a bit of research, Susan and I selected top of the line commercial-grade dry suits manufactured by Diving Unlimited International, DUI, in San Diego. These suits worked so well that the team of commercial divers we worked with in Canada wanted to know where they could buy them.
THE UNDERWATER HOUSING & CAMERA SYSTEM
The underwater housing problem was not so easily solved. Up to that time there had been little underwater work done in Showscan. Even though the new 65mm Cinema Products/Showscan camera was under development, all Showscan projects were still being shot with Panavision and Mitchell 65mm cameras with special motors to sustain the 60 fps rate. For my project, I used the very compact Panavision 65mm high-speed camera with 400' magazines because an underwater housing already existed for this camera/magazine combination, although it would need to be adapted to the needs of this project. Panavision had also designed a special motor and related control electronics package for this application, but it was one-of-a-kind and still had a few bugs. It was also a pellicle reflex camera and the magazines were the Mitchell type with a friction brake on the feed side and a belt and pulley on the take-up side. It was features like this that later gave us a great deal of trouble.
At the time, the range of 65mm lenses was severely limited and none of the available lenses was acceptable for this particular shoot. Fortunately, Barnaby was already researching large format still camera lenses in order to identify those acceptable for re-packaging in motion picture style housings. He was doing this as an ongoing project, to create a line of very high quality Showscan prime lenses to go with the new camera. One of the lenses that was on his list was a 21imm Mamiya, designed for 2 1/4" format cameras. Except for its slow speed, f 3.5, it was an extremely fine lens and was small enough to adapt to the 65mm Panavision camera in our underwater housing. In front of a 65mm target, it gave a 98 degree horizontal angle of view and had virtually no pincushion or barrel distortion. Now we had a great lens.
The next problem to overcome was to adapt the existing underwater housing to the needs of this project. It would have been far better to design one from scratch, but there was neither the time nor the budget to do so. The housing I had to work with had been originally designed to work merely as a "splash" housing. It was lashed to the deck of a coast guard surf rescue boat in a spectacular scene from an early Showscan production, New Magic. Later it was pressed into service as the housing for some actual underwater shots, but it was basically a massive box with a Plexiglas front plate for the lens to shoot through. There were some crude controls on the front plate for focus and aperture, and framing was accomplished via a crude, wire-frame sports finder. It also had an external box for the camera batteries. This crude housing had worked well enough for the surf-boat shots and some brief, shallow water shots of marine life, but I needed something more functional. I had to simulate the point-of-view of a three-ton, self-propelled, deep submersible. I had to develop a housing with more sophisticated features such as a dome port, buoyancy control, a video viewfinder and a propulsion system. It needed a lot of work.
We started with the dome port. In underwater work the most desirable lens port is really a dome shaped optic which cancels the magnification that occurs when light rays travel from water (outside the housing) to air (inside the housing) on the way to the lens and film. Lenses and dome ports need to be carefully matched. Designing and manufacturing the port to match the lens usually does this. It was another time-consuming process, which we had to find a way to shortcut. Barnaby found an off-the-shelf dome port, which was acceptable. We then built a test rig which allowed us to empirically determine, in water, the correct spacing between dome and lens, as well as lens back focus distance, which is also affected by the dome port. We did this by building a small darkroom over a test tank at Birns Oceanographic. Inside this dark room we mounted the dome on a plate which allowed us to place it face down in the water while keeping the backside dry. Attached to the back of the port was a jig, which held the lens, allowing it to be adjusted relative to the dome. Behind the lens was a ground-glass reticule whose distance (back focus) could also be adjusted relative to the lens. We then submerged a focus and linearity test chart under the test rig in the tank. Then, using a large magnifier, we were able to see the effect of our adjustments to the dome and back focus distances until we had an image that was flat and sharp from center to edge in the entire field of view. It also allowed us to determine the new focus calibrations for the lens. We then locked everything down in our jig and took it to the machinist who was able to take precise measurements from the jig and build the lens and dome port mounts to those specs. The test rig worked perfectly. It allowed us to test and perfect the optical system in a fraction of the time it would normally have taken. Later, we shot some critical focus, linearity and calibration tests from a similar rig built over my swimming pool. When those tests proved good, I then shot the first full underwater tests also in my pool. Everything worked perfectly and no further adjustments were needed. Next came the view-finding system. Reflex viewing would have been preferable but it was out of the question. It was impossible to re-design the housing to accommodate a reflex system. In addition, the pellicle reflexed Panavision viewfinder was far too dim to be useful in the low-light underwater environment. My next choice was a video tap on the viewfinder of the camera. However, there wasn't room in the housing for that either. So I finally settled on a parallax view finding system which utilized a miniature, black and white Sony chip camera which performed well in as little as 3 lux. Since the housing already had a Plexiglas front wall, I was able to mount the chip camera very close to the taking lens with a bracket, which allowed me to compensate for parallax at any distance I chose. The chip camera mounted an 8.8mm lens which gave me an angle of view of 41 degrees vertically and 53 degrees horizontally. The 24mm Mamiya gave me a field of view of 51 degrees vertically and 95 degrees horizontally. This allowed me to see 80% of the Showscan image in the vertical direction but only about 60% horizontally. However, since the "theatre" where this would be seen was designed to simulate the interior of Deep Rover, the images would be seen through the simulator's large, spherical view port, which strongly favored center screen composition. I just had to be careful to judge where the "unseen" left and right frame lines were so I could keep those areas of the frame free of unwanted material. Finally, to complete the video viewfinder system I put a 5" Panasonic monitor in a cylindrical housing with a bracket system that allowed me to attach it anywhere I wanted on the exterior of the main camera housing.
The final two steps in camera housing modification involved the design of buoyancy control and propulsion systems. Here there were two important considerations. First, we were going to be simulating the point of view of a 3-ton deep submersible. Because of its great mass, Deep Rover is quite stable in the water and that is how I wanted the camera to be. So trying to move the camera through the water by swimming was out of the question. Swimming might introduce oscillations which would spoil the illusion of being inside something massive and stable. The solution was to attach two Farallon propulsion units, one to either side of the housing, with fully adjustable brackets. These brackets allowed me to adjust the Farallon units to match the center of gravity of the whole assembly as well as to compensate for the irregular shape of the housing, which caused considerable asymmetrical drag in the water. Without these adjustments, the housing wanted to skew sideways and tilt downward as it moved through the water. Furthermore, deep rover had horizontal side thrusters that not only propelled it through the water fore and aft, but, when used independently or when run forward on one side and reverse on the other, cause Deep Rover to pirouette as delicately as a ballerina. Therefore, I modified the Farallons so that I could run them at variable speeds, forward and reverse. This gave me exactly the same control as the horizontal thrusters on Deep Rover. Second, Deep Rover ascends and descends in the water using vertical thrusters, which cause the neutrally buoyant vehicle to move up and down without pitching up, nor down. I found that I could easily cause the housing to ascend or descend just by pitching it up or down and using the Farallon thrusters. However, I needed a system to cause the camera to rise and descend without pitching up or down. To do this I designed a rigid buoyancy tank with its own compressed air bottle and supply and dump valves. This tank was attached to the top of the housing, directly over its center of gravity. If I wanted to ascend, I could press a valve to inject air and make it positive. If I wanted to descend, I could press another valve which dumped air, causing the housing to descend with around six lbs. of negative buoyancy. Finally, I attached a set of handles to the rear of the housing, centered on the thrust line, which had a control panel that allowed me complete control over the thrusters and buoyancy. The result was that I could fly the Showscan camera system through the water with great ease, although the entire system now weighed somewhere around 400 lbs., in air. Actually, its great mass helped to give the kind of stability needed for the simulation.
THE 'BIG DOME' PORT
There was one important element yet to be added to the housing in order to create the all-important opening shot of the simulation. As I've said, this entire film is from the point of view of someone inside Deep Rover, looking out through its huge dome shaped window. In the script, the opening shot described being lifted off the deck of a ship and lowered into the water for the beginning of the rescue operation. When you're inside Deep Rover being lowered into the water, you see the water rising slowly up the face of its dome window. That air-water interface line is in sharp focus. In our underwater housing, the lens port was too close to the front of the lens for that line to be anywhere near in-focus. We needed a way to create an air space in front of the lens that would push that interface three to four feet away, so the interface line would be sharp yet wouldn't vignette its 95% angle of view. The answer, of course, was an alternate dome port for the front of the housing, for that one shot.
So we acquired a four foot diameter plastic dome of optical quality and designed a large aluminum plate on which to mount the dome with a watertight seal. We then cut a hole in the center of that plate to accommodate the lens and worked out a way to attach the whole dome/plate assembly directly to the front of the underwater housing. This required that we remove the small dome port, attach the aluminum plate to the front of the housing with a series of bolts, and then attach the big dome to the plate with another series of bolts. The whole process took about an hour if things went well. Furthermore, the volume of the big dome required that we add about 1200 lbs. of ballast to counteract its buoyancy. We also needed a system to blow a curtain of bubbles in front of the lens at the end of the shot. This was to simulate the blowing of ballast by Deep Rover and would be our transition device from shot to shot throughout the simulation. Thus, when we finally did the opening shot where we were lifted off the deck and lowered over the side into the water, the total weight of the camera system - including the bubble system with its air tanks and valves, myself, an assistant, an effects man, a grip, the platform we were mounted on and all our ballast - was around 1.5 tons.
MAKING THE SHOTS
Ultimately the system gave us what we wanted but it gave us plenty of headaches too. For starters, the opening shot required that while we're being lowered into the water we see the deck crew on our ship, a television news crew reporting on the rescue operation, a rescue helicopter hovering in the middle ground, a Coast Guard ship standing-by in the background, a tow vessel for the robot sub Manta making a pass in the middle ground, and Deep Rover and some scuba divers in the foreground. At Showscan speeds, our 400' magazines of 65mm negative lasted 1 min. and 15 sec., after allowing 50' for slate, run-up and run-down. On dry land, this kind of coordination and time limitation would have been difficult, but manageable. In this situation, it was a nightmare.
As I've mentioned it took about an hour to remove and replace the big dome which was used for the opening shot. As scripted, the opening shot took about a minute to play out. This meant we had one take per load with an hour to re-load and re-set all the vessels, the actors and the helicopter. And, of course, we had an unanticipated problem. The water temperature was around 40 degrees, but above water it was clear and sunny with an air temperature of around 70 degrees. Thus, while setting up for a shot, the big dome and the air inside it would heat considerably. When we then lowered it into the water, the sudden cooling caused a heavy moisture condensation on the inside which completely fogged the dome, making shooting impossible. Finally, with a combination of working in the shade, large quantities of silica gel, special anti-fog chemicals and a spray of water to keep the dome cool, we managed to overcome that problem. In the end, it took six takes and a whole day to do that one shot.
In addition, there were more problems. The 65mm Panavision high-speed camera was not designed to run continuously at 60 frames per second. At that speed the large, thin, glass pellicle vibrated tremendously. It was fortunate that we had our crude, parallax video viewfinder system because the image from the pellicle was unusable due to the vibration. Removing the pellicle was not an option, however, because we needed it to verify the alignment and calibration of the video parallax system for different focus settings. Moreover, we couldn't remove it for shooting because that would have changed back-focus, which was very critical due to the wide-angle Mamiya lens. Therefore, we did our best to live with it, but it wasn't pretty. At the end of each take, checking the gate for a broken pellicle was as important as checking it for hair. And it did break -- several times.
We also had problems with power transistors burning out in the motor-speed control circuits of the camera. This was due to continuous running at 60 frames per second with equipment that was not designed to do that. And we were plagued with film jams and breaks due to not having proper controls for ramping the camera up to speed and down again. We had friction braking on the feed side of the magazines and a belt and pulley on the take-up. With that type of system it's bad enough shooting 65mm at 24 frames per second, even when you can put your hands on the feed and take-up pulley to keep proper tension on the film. With the whole rig in the housing, we couldn't adjust tension at all during takes. And we couldn't take up slack at the end of a take if the feed side rotated a little too far after cutting. We were able to fine-tune the feed-side braking system to work semi-reliably. But it was usually a re-load if we started a take at the head of a roll and then had to shut down quickly. The inertia of a 400' roll of 65mm negative was just too much if the shut down came early; it would leave a lot of slack film in the magazine. If we attempted to re-start in that condition it would break the film every time. If we set the brake tight enough to stop that massive roll of film from turning after a 60 frame per second shut down, it was too tight for the transport system and tore perforations. This was especially true near the ends of rolls. We had similar problems on the take-up side too.
READY TO GO
Finally, on July 8th, we shot extensive open-ocean tests off Catalina Island. Everything worked well and all was in readiness. On July 19th, we departed for Canada and on the morning of July 22nd a small army of people and equipment descended on the remote location of Port Alexander, some 30 miles out of Port Hardy near Balaklava Island. And there we sat. As luck would have it, the window of opportunity (the statistical likelihood that we would have good water clarity and good weather at the same time) not only didn't open, it was nailed shut! The water clarity at that time was the worst it had been for many years. An unusual warm spell had caused a plankton bloom which rendered the water thick with green algae; so, we waited. Hoping that the water would clear, we spent the first three days in further testing and preparation. There was plenty to do with respect to finding the correct bottom placement for Sea Otter, and there was a lot of rehearsing and coordinating of the commercial divers and my Showscan camera crew. Finally, on the 25th, we decided to do the very difficult opening shot where we were lifted off the deck of the Redonda and lowered into the water. In fact, the good weather and soupy water helped us for that one shot. The water was calm and this facilitated the complex coordination of ships, divers, helicopter and subs. In addition, the murky water actually made our visual transition to the next shot a little easier. However, for any other shooting, it was a bust so we pulled the plug. On July 30 our "army" was on its way back to Vancouver and Los Angeles and I was on my way to Halifax, Nova Scotia with Doug Trumbull. We needed to consult on one of the other Showscan projects that was being shot for Expo 86. Now that the predicted "window' had failed to open we were faced with the only other possible choice. We would have to wait until October when the water would certainly be much clearer, but the weather would be more difficult. Not only would we have to deal with frequent, almost constant overcast, which would greatly reduce underwater light levels, we would also have winter storms which could seriously jeopardize people and equipment.
TIME TO RE-GROUP & DO IT AGAIN
The month of August and the first half of September were spent further refining the underwater camera system as well as working on some other Showscan projects. By September 16th, I was back in Vancouver for more scouting and checking of weather and water conditions. On the 18th, Susan Walsh and a couple of the Canadian divers joined me. We spent three days on our camera boat, Sea Tech, shooting coverage which would be used to get us from the opening shot to the point where we're approaching the disabled sub, Sea Otter.
One of the best moments was when we stumbled into a massive school of jellyfish in Pat Bay, near Victoria BC. We were doing "fly-by" shots of the surface-towed sub, Manta, when we encountered tens of thousands of the creatures in an area of several square miles. We were able to do some great shots of Manta, as she gently glided through this galaxy of delicate, white, star-like creatures. By the 22nd, we were returning to Los Angeles with another couple of shots in the can.
On October 2nd, I flew back to Vancouver and on the morning of the 3rd, diver Lou Lehman and I donned our dry suits and loaded our tanks and other diving gear into a Cessna 208 floatplane. We spent a long day flying north, from Vancouver to Port Hardy, landing directly on our location and diving on it from the floatplane for a final check of water conditions. We also landed and dove on numerous, promising back-up locations. We did this as a hedge against further problems on the primary location. However, the weather and water conditions were looking good everywhere and I was growing anxious to get started again.
Finally, the decision was made: it would be Port Hardy and our original location, Port Alexander, again. On October 4th, we loaded the support ship Redonda for the second time. Its decks were again crowded with subs, generators, ATCO trailers for control rooms and equipment storage, and endless cases of diving and camera equipment. Redonda then sailed for Port Hardy, while my crew and I began the journey by car and ferryboat.
By October 7th, all crew and equipment were on location at Port Alexander and preparations for shooting were begun. The following notes from the daily production report give the best account of how the shoot went:
October 7: 0700 setting up on location at Hussard Point/Port Alexander. Redonda arrives location at 0300. Sea Tech arrives Port Hardy at 0700. Depart 0800 and arrive location at 1000. Very good weather conditions, sunny, calm sea. Underwater visibility over 50ft. Ambient light levels at working depth are 256 foot candles. Adjustments made to camera, balancing weights and light bar. Made test dives with full lights and co-ordination of support divers. C an-Dive setting up Deep Rover and replacing damaged parts on Sea Otter at 1800. Main crew wrap at 1830. Hysub crew arrived on 1630 flight at Port Hardy; to location and set up of Hysub till 2145.
October 8: 0700. Security meeting for all crewmembers aboard Redonda. Camera crew made final balancing tests and reported ready to shoot at 0930. Delays due to very complicated procedures in setting up the three submersibles on bottom of cove at 30 feet. Leak in hydraulic system of Hysub. Spare part ordered from Vancouver. Underwater welding test completed. First shot completed at 1642. Also video shots from Hysub. Weather conditions sunny in morning, overcast in afternoon. Calm sea. Underwater visibility more than 50ft. Ambient light levels drop to 125 fc. Wrap at 1830.
October 9: 0700. Gusty winds between 20 and 35 knots. Rainy rippled sea. Underwater visibility still over 50ft. Ambient light level on bottom at 128 fc. All subs in position at 1100. Valve on diving suit of cameraman C. Barbee broke. Backup valve substituted and replacement ordered from DUI, San Diego. Camera ready at 1100. First shot at 1224. Box lunch at 1300. Due to strong swells, subs are constantly changing position. Unable to get shot of cutting cable. At 1424 decision to stop photography and pull subs up. At 1308 all vessels recovered. Wind increasing to gale force. Eight foot seas in Goletas Channel. No vessels from Port Hardy able to reach location. At 1711 Coast Guard vessel Bull Harbor on way from Port Hardy with blankets and additional food for crew. Decided to spend night on location and expect weather improvement as per forecast. Crew had dinner at 1830 and to sleep on board the four vessels: Redonda, Sea Tech, Stel Star and Oceaner. Shots completed: Camera approaching Hysub, Deep Rover and Sea Otter together. Not completed: Underwater cutting-torch sequence and video shots.
October 10: 0730. Weather forecast at 0545, high-pressure building west of Vancouver Island. Sea rippled at 0800. At 0900 decision to dive location and scout visibility. Vis ok, but moderate. Hysub launched and crew starts as scheduled. At 1200, no improvement on weather condition. Continue filming with Hysub, but leaving other subs on board Redonda. At 1300, weather deteriorates. Winds picking up to 35 knots with eight foot seas. At 1400 shots with Hysub completed. Included video shots of Hysub grabbing cable with manipulator. Wrap and secure subs and camera on boats. Oceaner leaves Port Hardy at 1445. At 1546, Stel Star radios that ship will stay over night in Christie Passage. Sea Tech radios she's ok. On way to Port Hardy. 1630 Redonda weighs anchor at location. Heading for Port Hardy. Gusty winds from SE. Seas at eight feet. Arrive Port Hardy at 1906. Shots completed: Hysub passing camera, video shots.
October 11: 0600. Based on weather reports expecting 48 hours calm sea. Decision to continue on primary location. Redonda leaving Port Hardy at 0630 to location. At 1000 Hysub inoperable due to leak in hydraulic system. Decided to continue welding sequence without Hysub. Shots on cutting cable are completed. Also one version of Sea Otter free and rising to surface.
October 12: 0600. Weather forecast winds up to 35 knots from SE for Saturday. Camera and vessels ready at 0930. Breakdown of generator on Redonda delays launch of subs. Sea Tech and Rye Whiskey leave with divers and camera to "daisy-field" for fly-by shots. Generator problem located at 1100. Camera reports problem with video viewfinder. Test run of generator at 1300 looks ok. At 1526 underwater welding and cutting shots completed. Wrap at 1645. Sea Tech leaves location at 1756. Redonda at 1830.
October 13: 0630. Filming in Beaver Harbor near Port Hardy for shots with Deep Rover reaching bottom. Also 16mm Hysub pov shots of Sea Otter sub driver through portholes for simulated Hysub video feed. Shooting completed at 1900 and return to Port Hardy. Wrap dinner for all crew and seamen at the Sea Gate hotel.
October 14: 0900. Redonda and subs in transit to Browning Passage for filming on wreck of Themis. Set up with nine support divers at wreck at 1651. Wanted shots of Deep Rover over wreck in 80 feet of water. Surface conditions and currents too dangerous to place deep rover over the wreck. Shot abandoned. Wrapped location at 1756 and return to Port Hardy.
And with that, the shoot was finished. The project had lasted from May through October of 1985, but the finished film was only 4 minutes 45 seconds long. In terms of sheer physical and mental difficulty this was, without doubt, the most demanding and difficult project of my career. There had been enormous technical obstacles to overcome, but we did overcome them. In the final 8 days, we shot 95% of the total project. We struggled against Mother Nature and the laws of physics; battled near freezing water, fatigue, frustration, equipment failures and gale force storms; but we got what we came for.
In its final form the Deep Rover simulator "ride" was quite a hit at Expo 86. The Canadians had what they wanted: An interesting and fun way to show their technical prowess in the field of undersea exploration, construction and rescue. At the Expo, people could enter the simulator capsule in shirtsleeve comfort, be lifted off the deck of the Redonda and be lowered into the water. They then descend to a great depth to watch the real Deep Rover and its companion remote subs cut through the cable that held Sea Otter and her crew captive in the icy darkness. Through this 'simulated' experience, audiences could get some sense of what it would be like to actually do such a thing; but they would never know what we went through to create the illusion.
Last Update: March 26, 2017 Web Author: Chuck Barbee
Copyright ©1999 Charles L. Barbee - ALL RIGHTS RESERVED