With Sanctum opening this Friday in IMAX digital and other digital 3D theaters, the Kinotech Blog is taking a retro approach to explore the origins of James Cameron’s underwater 3D films. In 2003, I sat just a few feet from Cameron as he explained his history in working with IMAX and the creation of the Pace 3D rig. It was here, during his keynote to the Large Format Cinema Association, that Cameron spoke about his collaborations with giant screen filmmaking legend Stephen Low (Cameron would act as Executive Producer on Low’s Volcanoes of the Deep Sea) and his extensive work with David Keighley on post production. I’ve seen Cameron speak a number of times since, including the big DLP 3D demonstration at ShoWest, where he was one of five major directors endorsing the technology, but it was here at the LFCA, in a room full of filmmakers and post production and vfx professionals, that I found him most accessible.
One further note – during the Q&A after his keynote, I asked him if Expedition: Bismark, his Discovery Channel documentary chronicling a dive to the famed German battleship, had been filmed in 3D. He confirmed that underwater sequences had. Here’s hoping it will air as shot on 3NET soon.
James Cameron’s keynote, titled “Adventures in the Giant Screen Trade,” was presented during the Large Format Cinema Association’s (LFCA) annual conference May 14-16, 2003 in Los Angeles. The LFCA was kind enough to give me a transcript of the speech, for publication in the July/August 2003 issue of The SFC Review. It is reprinted here with the permission of the LFCA’s successor, the Giant Screen Cinema Association (GSCA). Special thanks to both Jeannie Moore and Tammy Seldon.
I titled this address “Adventures in the Giant Screen Trade, or What I Did On My Summer Vacation.”
So I’ve done one, count ‘em one, title in large format, and I get to do the keynote address at LFCA.
I assume this is because it is a curiosity whenever a Hollywood film-maker crosses over to work in large format, so I’ll give you my impressions, having done it once, for what they’re worth…
…but bear in mind I didn’t do a drama, it was a documentary which doesn’t speak well to my core skills.
However, we plunged in heedlessly and I learned a lot about documentaries…like: they’re hard. Major props to documentary film-makers…I have much greater respect now for the rigorous discipline of non-fiction than I did before, believe me.
I’m going on the assumption that everybody here today is already a fan of, or practitioner in, giant screen filmmaking. So there’s no point extolling the virtues of the large format experience,which I’ve loved since I saw Graeme Ferguson’s at Ontario Place in 1972. I’ve always wanted to make one of these films, and much more so since Imax 3D was introduced.
I got my first taste of shooting when I co-directed for Universal Studios. It was a location based 3D film which we shot in 5 perf 70mm, and it was projected using dual 70 projectors running at 30 frames per second.
The goal was to do lighting, camera movement and editing as close to the style of the films as possible. I was told going in about all the limitations of stereo photography, so I got Universal to pay for extensive testing to find out what worked and what didn’t.
We consulted with all the gurus of stereography and was told what I could do and couldn’t, and we looked at all the existing large format and location based 3D films to make notes about what worked and what didn’t.
We found for example that it took several seconds for the eyes to adjust when you cut from a distant subject to a close one, so we knew that the normal editing rhythm wouldn’t work.
Then we started shooting and I was amazed at how bulky and difficult the cameras were.
Big beamsplitter rigs the size of a refrigerator requiring night exteriors to be lit to an F 5.6. When we did our big future war set, we used 3 Muscos, 3 condors and 1Beebee Light.
I don’t know how anybody else in Hollywood was making a movie that week…because we had all the lights. The gurus told me I should expect to get 2 setups a day. I told them I would do 10. It turned out to be more like 4, but I still consider that a victory.
Despite a rocky first date, I was still in love with 3D. But I was dissatisfied with the limitations of the cameras…the size, weight, the reload time, the high F stops, the lack of a zoom lens and active convergence, and the inability to shoot good production sound because they were so loud.
Stephen Low and I got talking about high definition, because he had seen some good results shot in the deep ocean.
We wondered if maybe, just maybe, shooting two HD cameras side by side could produce enough picture information, when the images were combined in 3D projection, to fool the brain into thinking it was seeing some approximation of Imax 3D.
So we co-financed a set of tests, which were shot up at Imax headquarters in Toronto. We shot with two Sony F-900 cameras mounted on a beamsplitter rig, and also with two H-10 cameras mounted in parallel on a plate.
Each subject was also shot with an Imax 3D camera. We made some selects from the HD material and David Keighley blew it up to 1570 for us, then we projected it along with the Imax 3D for a battle of the bands.
The results were interesting. The HD material was not as well resolved as the Imax, as you’d expect, and it was of course a different aspect ratio…16 x 9…but it looked surprisingly good.
I was really encouraged, so I went to Tokyo to meet with the Sony HD engineers at their Atsugi facility, where the professional products are designed…and I proposed that they build a special version of their 900 series 24P cameras for 3D photography.
They would need to physically re-package the cameras so two of them could fit within the average human interocular distance…the distance between the pupils of the eyes…which is about 69 mm.
During the 6 months it took them to build the prototypes, Vince Pace, of Pace Technologies, here in LA, worked on the systems integration. He built a motion base and a servo control system, so the cameras could be actively converged on the fly, and a lens control system.
Panavision agreed to build some special prime lenses for us, again they had to fit within the 69mm constraint, and Fujinon did the same thing, building 3 matched sets of 10-1 zooms for us which had custom reduced front elements.
6 months later the prototype cameras arrived from Japan and Vince put the whole thing together and fired it up.
We connected the camera system to digital projectors at my office and had a great time playing around with stereo images in real time.
That first phase of testing, spending hours looking at what worked and what didn’t on a movie-sized screen, taught us more than all the books on stereo photography we had ever read.
Vince and I concluded that the hot tip was to slave the convergence to the focus.
The idea was that, as a cinematographer, you tended to put your focus on the subject of greatest interest, and that in turn was where the audience as most likely to be looking…so just put the convergence there also and their eyes won’t have to work to do the fusion.
The theory was if you were always converged where the audience was most likely to be looking, then there would be no eyestrain, and no adjustment time from cut to cut.
If that proved to be true, then we could cut in a normal movie cutting rhythm.
We thought we were on to something, and Vince developed the system so that the software automatically placed the convergence plane at the plane of focus.
Next it was time to test the camera system in the real world. Vince suggested going to a park and shooting a few brightly colored objects. I suggested getting together my friends who own classic World Wear 2 aircraft and doing simulated air combat.
So the first shot done in the field with the new system was me hand-holding in the jump seat of a P-51 Mustang carving 3 G turns over a rice field outside Sacramento.
Then we crawled around hand-held inside a B-24 Liberator in flight, firing the 50 cal waist guns, and did some air to air stuff with a Tyler major mount in an A-Star helicopter, and then did some dolly mounted shots on the ground at sunset.
The camera weighed 24 pounds including the two zoom lenses and matte box, and was very easy to use, just like any production camera I would use on a movie, except it ran for 2 hours and took one minute to reload.
It was a remarkable day because we shot hours of exciting material, on several different mounts and in different difficult environments…probably twenty five or thirty setups. This was light years beyond what I had experienced on , on every level.
So we were very revved up and excited, and started trying to figure out what our first project was going to be.
I hadn’t gotten a script I liked yet on the Mars film, so it got pushed back. We needed another subject.
Meanwhile, my brother Mike had been building these two little robot vehicles to explore the inside of Titanic. After three years of development he was nearing completion, so it occurred to us that we could do a 3D Imax film about exploring the Titanic wreck.
3D HD in the deep ocean…never been done.
And if I couldn’t raise money for a documentary about Titanic, I should forget about documentaries.
From the time of that decision to the time we were on a Russian research ship heading for the Titanic wreck site was less than 6 months.
In that time we had to raise the money, build a deep ocean pressure housing for the camera, adapt the optics to the dome port of the housing, and do the electronic systems integration to mount the whole thing on the sub. None of which was easy, or fast.
Also in that time we built a 3 ton lighting rig called Medusa, which was deployed from a second ship over the wreck, and we also prepared our little ROVs, which were named Jake and Elwood, for integration to the Mir submersibles so we could launch and recover them at the wreck.
The Titanic wreck lies two and half miles down in the North Atlantic. To give you a sense of what that means, it’s like looking down at the Queen Mary from an airplane flying at 12,000 feet…it looks like a grain of rice. If you could strip away the water, that’s what Titanic would look like from the surface.
So you are going to free-fall for 2 hours in a 3 person submersible, most of that time in pitch blackness where no sunlight has ever penetrated.
At the bottom the water pressure is 5,500 pounds per square inch, so your camera housing has to withstand that pressure. And you have to light up this giant wreck with lights you’ve brought with you, because there is zero ambient light.
Mike Cameron built the titanium housing for the camera, and Vince Pace did the optical integration, which involved grinding a special corrective diopter to adjust the fields of view of the two lenses to the single dome port of the housing.
The housing was the largest implodable volume ever dived on a manned submersible, and if it failed structurally the force of the implosion would be like a depth charge, and would kill us in the sub…so the engineering specification on the housing was very high.
Essentially it was built as if it was a man-rated submersible itself.
It was kind of crazy to layer all these new and untried technologies on top of each other to do this expedition, and then to shoot the whole thing in the new format of HD 3D for an Imax market.
Looking back it was ridiculously risky, not from a safety standpoint, but from the perspective of having any of it work.
But it did work. It was frantic, and we made a lot mistakes at first, but we figured out how to shoot the 3D HD on the ship, and to document the expedition day by day.
More importantly the deep camera system worked like a champ.
I’ll never forget projecting the footage from the first dive in 3D when we got back on the ship. It was like a religious experience. There was Titanic, lit up by 16,000 watts of HMI light, looking so close you could reach out and touch it.
By having the camera outside the sub on a pan tilt, we were able to do film-style angles and moves, like looking down into the hatches or up along the cliff-like hull…stuff you could never do shooting through the port-hole of the sub. On the first dive we got more usable shots than on all 12 dives when we were shooting on film during our ’95 expedition.
And they were all in 3D.
All the problems and headaches and ulcers of the prep period melted away. It had all been worth it.
Over the next few weeks we continued to shoot the wreck, including diving our two bots deep into the interior to see things which hadn’t been seen since the ship sand in 1912.
But more importantly we were working out how to shoot with the new camera system. Mounting it on a crane on a remote head, or putting it on a steadicam, or shooting hand-held out in a zodiac in a storm.
By the time we were done with the expedition we had shot over 300 hours of Imax-compatible 3D, including helicopter aerials, boat mounts, dolly shots, steadicam, handheld, remote crane, and shallow and deep underwater shots.
The next challenge was post production.
The good news was we had 300 hours of material. The bad news was we were making a one hour film.
There had been no script, no shotlists, no directing in the Hollywood sense. No second takes of anything. No action and cut. Just endless hours of people doing what they were doing during the expedition.
It was a true documentary, which meant it was going to get made into a story in post.
Trying to structure all this material into a narrative was one of the toughest creative challenges I’ve ever faced as a film-maker.
We were telling a story of an expedition, with it’s people and discoveries, but also telling the story of Titanic with her people and their true stories.
Weaving those two together, in a compelling way was daunting.
It took a year to put it all together, and to do the visual effects which bring the wreck to life.
We learned an awful lot about making stereo during that time, and a lot was possible in digital post production that would not have been otherwise.
For example, we found that it was relatively easy to do digital post-convergence of the images, once we got the hang of it. And it turns out that convergence is really critical on the Imax screen, which is much less forgiving than a normal-sized movie screen.
The final hurdle remained.
We were making a film for the Imax 3D market, but it had been shot in high def.
Once the digital master of the film was done, all the effects inand the color timing complete…we had to get it out to 1570 film.
I held my breath during this process, because if it didn’t work we were going to look like a bunch of dorks.
We looked at film-out tests from a number of companies, including CFI, EFILM, and nWave, and they all looked great. CFI wound up doing it, and they did a fantastic job for us on a tight deadline, with the assistance of David Keighley.
Though we had DLP projection at our production facility, which we were using as we went along to judge our image quality and visual effects…when we went to Imax we saw a whole bunch of new details with a clarity we had never seen before.
This was good and bad.
A few underwater shots had electronic interference from the sub’s hydraulics which we hadn’t seen before. We were able to swap them out or clean them up with noise reduction techniques, but it was surprising to see how much more detail was visible than we had ever seen before.
I think this is significant because people usually judge HD by looking at it on HD monitors or projectors, or at the most by filming it out to 35mm, and the simple fact is that none of these display methods show all the detail which is captured by these phenomenal cameras.
The Sony 24P Cine Alta cameras are capturing at a much higher resolution than 35mm negative, even daylight negative at 50 ASA. I would say it is about the equivalent of 65mm 5 perf negative. People have told me for the last two years that I’m nuts for saying this, but it’s true.
At any rate is in theaters now, so you can judge for yourselves.
I’m not saying HD equals Imax, and I don’t recommend it as a replacement for 2D Imax, but it certainly seems to be working for audiences as an alternative for 3D Imax.
People tend to get overly fixated on numerical analysis of resolution…pixel counting.
But there are a lot of other issues. For example, the size and weight of the camera determines the kinds of shots you can do. And the kinds of shots in a film determine its style, its pace, and the way in which it tells a story.
The dynamic possibilities of an HD based camera system for large format I think vastly outweigh the slightly lower resolution. This is a trade off I was willing to make, and I suspect other film-makers as well.
And another thing to consider is that digital cameras are improving all the time. I saw the new Sony 950 cameras in Tokyo a couple of weeks ago and they are stunning, with more dynamic range and improved colorspace. But beyond that, the possibilities for improvement in digital capture are still open ended, whereas film doesn’t have much headroom left.
The large format business will die without new content, and that content has to come from new film-makers coming into the field, as well as those already practicing.
To attract film-makers, especially those who will bring vigor from other disciplines, the camera and post production tools have to be familiar to them.
They have to be able to shoot and edit and do effects the way they are used to doing them. The digital tools offer this, and hopefully will bring about a renaissance of 3D large format production.
I would encourage everyone here to continue to develop tools and methods which can bring new energy to the large format production world, both on the technical side and the business side.
The funding and distribution paradigms need to be looked at. As long as film-makers need to make deals theatre by theatre the business will never survive.
The projectors are too expensive and we need a lot more screens,so we need a technical breakthrough there. Maybe it’s a 4K chip. Digital is probably the answer, it seems to be the answer to everything else.
We need to continue a tradition of showmanship, of razzle dazzle, by whatever means necessary. 3D falls into this category, and I would encourage large format film-makers to rely less on screen size and more on the inherent fascination which viewers have with the illusion of depth.
And we need better marketing strategies. Walden Media produced , and Disney did the distribution, and they pioneered some new techniques.
They did a fantastic multi-level marketing and presales campaign which used conventional 35mm theatrical trailers, 3D and 2D trailers, regional and local TV spots and print media.
They also did a very successful outreach program to schools, sending out an educators’guide which showed teachers how the film fit their curriculum goals in social studies, science, and history.
Disney used their 1-800 line operators for group sales bookings, and over two million dollars in tickets were sold even before the film opened.
It’s this kind of thinking which can turn this business around and make it sparkle again.
As Stephen Low tried to warn me, this is not a wildly profitable business.
We do this because we love it, and because we can make films for this market which are unique and epic and exciting.
We do this because we love to put on a big show, to transport people to worlds and places and experiences which they could never have in real life.
As John F Kennedy said when he launched the Apollo program: “We do this not because it is easy, but because it is hard.”
Well, these films are damned hard. But they’re worth it. And God bless the folks with the tenacity to make them.