by Gord Harris
What are the realities of the 3D LF world right now, and what is coming up next? I have always liked the mission of Roland Music: “We create the future.” It suggests an attitude of responsibility and risk-taking rather than just responding to market demands in a reactive fashion. While the early growth of the 3D theater network was almost exponential up to 2001, the boom is over, and 3D theater sales have stagnated dramatically. But if you review the box office figures, 3D films have proven consistently popular. Unfortunately, many LF filmmakers today declare that it is impossible to make a return on investment shooting a 3D LF film given the current economic model. So what technology factors might change this?
Many believe that the 3D LF theater network will grow only if we stimulate the production of more and better 3D film product. So how might that best be done? One scenario is the evolution of several alternatives to coexist with the present options of shooting in 15/70 or 8/70 film. These should make it faster, cheaper, and easier to make good 3D LF movies. Some solutions are more likely to be used for documentaries, others for dramatic storytelling for fictional films. Let us consider some of the options:
In 1998 Marty Mueller of MSM Design designed a single strip 30-perf 3D camera that captured left- and right-eye images on one strip of film. The camera was intended for shooting 3D in the Space Station for Toni Myers and Graeme Ferguson of the IMAX Space team.
His brilliant design addresses most of the shortcomings of the earlier 3D rigs and dual strip IMAX 3D cameras. The 3D30 is approximately the same size and weight as a regular 15-perf 2D camera. It uses a normal human interocular spacing of 2.5” and is extremely fast to load by virtue of a pre-formed, drop-on film loop in the one-piece coax magazine. The prototype cameras performed extremely well and reliably in the course of filming Space Station. They captured shots in space that would have been simply impossible to get with any other 3D film camera. And as Marty Mueller says, “I think the best bang for the buck is film.” The simple, compact 3D30 will make 3D film production on the ground faster and ultimately cheaper in coming years.
On the down side, using the 30-perf camera is not cheaper yet because of the high post-production costs and delays of optically separating the left and right eyes so they can be shown on regular dual-strip 3D projectors. A special optical printer design by MSM is currently on hold due to “chicken and egg” film demand and supply issues. Once this 3D30 printer is finished, costs will come down dramatically, but it will take vision, generosity, imagination, and courage by film producers to do so.
When shooting with the 30-perf camera, film loads last half as long but cost just as much as shooting dual-strip 15/70. But for LF documentary filmmakers like Toni Myers, Greg MacGillivray, David Douglas, James Neihouse, and Bayley Silleck who insist on the highest possible image quality, this 3D30 camera is a great choice. It opens up possibilities for more involving and exciting 3D films, thanks to its smaller footprint and profile, easier reloads, and ability to be carried by one person. It also fits more existing camera mounts and heads, requires smaller accessories, and lowers shipping costs. It truly makes 3D production faster and easier.
Another option is to shoot 3D LF movies “digitally,” which actually is just the trendy hype name for filming them on videotape, albeit high-definition video. HDCAM tape costs about $100 for 45 minutes and requires no processing, as compared to tens of thousands of dollars for 15/70 film.
The most commonly used camera is the Sony 24p HDW-F900 camcorder, although Panasonic, Thomson, and others also make high definition video cameras. At present these use 1920 x 1080 pixel CCD’s, but compress the image by about 5:1 to record them on tape, reducing the effective luminance resolution to about 1440 horizontal pixels by 1080 lines vertically. Clearly, that is much less than film.
If you have seen Star Wars: Episode II – Attack of the Clones projected digitally or in 35mm theaters then you have seen what these cameras are capable of — and not capable of! Soon we should also get to see James Cameron’s Ghosts of the Abyss, which was also shot entirely in 24p. His Titanic expertise, big name draw, and passion for the subject bodes well for a wide release with commercial LF operators, who should also benefit from the marketing muscle of Disney behind the distribution and advertising.
You may have also seen some of the limits of HD-to-15/70 conversion in tests shown at recent LFCA and GSTA conferences. After doing some interlaced 2D HD video-to-film conversion tests for Stephen Low in early 2000, Imax Corporation agreed to do some joint tests with Sony, Low, and Cameron in May 2000 with the newly available 24p cameras.
We devised three separate 3D shooting rigs: a beamsplitter over-under rig, a side-by-side F900 mount, and a handheld HC10 industrial camera version for Steadicam use. We then shot comparison shots with the conventional dual-strip IMAX 3D Solido camera. The digital material was output on digital film recorders at DKP/70MM, Inc., and compared to the film version.
The results were encouraging. While the resolution was obviously lacking on wide shots, close-ups fared slightly better. The extra depth of field and shadow detail of the video cameras was good enough that Cameron decided to shoot Ghosts of the Abyss on HD video. He proceeded to work with Sony and others to produce a special lightweight head for filming. We can all look forward to seeing the 3D results soon!
Filmmakers like George Lucas and James Cameron are more concerned with compelling stories, great characters, and imaginative images then veracity or ultimate image quality. They prefer to shoot many takes of dramatic scenes that depend heavily on complex human performances and special digital effects. Because Lucas’ films are one big digital effects composite, he can afford some compromises in picture quality for the sake of digital control, cost, and convenience.
Lucas helped push Sony and Panavision into creating practical 24-fps HD cameras and lenses for Star Wars: Episode II, which contains 2,100 visual effects shots. Shooting on HD video saves film stock costs and scanning costs, and allows the director and crew to judge performances on set with 36-inch plasma video monitors. For 3D an added benefit of HD is the extra depth of field. Lucas loves “the malleability, the flexibility of this digital medium in which I can create certain characters, sets and locations that I can’t do any other way”.
On the down side, it is impossible to shoot slow motion with 24p cameras, and the lenses have to be about 2.5 times as sharp as 35mm lenses. The linear CCD’s do not handle overexposed highlights well and this limits dynamic range of contrast in shooting, particularly outside, where the lighting can’t be completely controlled. The cameras have many menu settings which if misadjusted can cause problems and inconsistencies. For Star Wars: Episode II the effective vertical resolution was only 800 lines, since a 2.40:1 anamorphic widescreen “scope” aspect ratio had to be extracted from the 1.78:1 (16/9) format.
In the future, better sensors, direct-to-disk lossless uncompressed image capture, and Quad HD resolution cameras will improve the resolution, dynamic range, and color look of digital, and give film a real run for its money. At the GSTA conference in Toronto, Chris Reyna showed a nice demo of a prototype 4K Olympus/NHK camera with about 8 megapixels, four times higher than current HDTV systems.
Other manufacturers are also hard at work at digital cameras that will be a vast improvement over HDTV for LF use. At present these are limited by the huge RAID hard drive systems needed to store the terabytes of information. By comparison, film is still cheaper, more robust and far more compact. But stay tuned…
One very promising approach is to shoot with smaller-format 3D film cameras and then optically or digitally print to LF film for projection. Filmmaker Stephen Low, stereographer Sean Philips, DP Bill Reeve, and others have investigated designs of much smaller 3D cameras based on VistaVision 8 perf 35 film or similar formats. Good 3D work has also been done in 5/70 by pioneers such as Colin Low and Ernie McNabb of the National Film Board of Canada and Murray Lerner for Disney at EPCOT.
A side benefit of shooting 3D in smaller formats is that smaller, quieter cameras are less disruptive to actors. Val Kilmer reportedly asked, “How can you act in front of a chainsaw?” after his first experiences on Wings of Courage with the big 3D LF rigs and dual-strip cameras. The 24p HD cameras and Arri’s 765 blimped 5/70 cameras are capable of filming virtually silently.
However, like HD, the cost of scanning, converting, and recording or optically printing smaller formats to LF is still rather expensive today. Nevertheless, the idea of using smaller 3D film cameras is sure to appeal to some, if not all, LF filmmakers. We will see more of this.
An option being tested by Disney, Imax, Imagica, and others is to repurpose conventional 2D 35mm films. These films usually cost many times more than typical LF films and hence have major production values. They might be Hollywood feature films with top stars and directors, or animated CGI films like Imax’s Cyberworld. nWave Pictures has also done well by skillfully re-purposing ride films and simulator attractions into films like Encounter in the Third Dimension.
Digital image processing techniques and algorithms have advanced enough such that quite convincing 3D can also be generated from 2D film or video images. Samuel Zhou’s DMR™ team at Imax gave a recent example in Space Station with the launch shot from inside the shuttle, which was shot in standard NTSC digital video. Other examples, such as a Super Speedway test, have been shown at industry conferences.
Imax’s DMR technology continues to develop. It is now possible to produce good quality results on 15/70 film from much smaller film frames. For example, Apollo 13: The IMAX Experience truly worked (for me at least) as a dramatic story, despite the fact it was originally filmed in 35mm.
Minor deficiencies in LF image quality are not as problematic in truly engaging story films, where you are sucked in emotionally by great acting and a great plot. As the quality and cost of scanning, recording, and rendering improves, sooner or later an entire feature shot in 2D on 35mm or HD will be converted into stereoscopic 3D LF and successfully released in commercial venues. I know that emotions run high in our industry as to whether repurposing is a good thing, but 2D to 3D conversion is a technology to watch. It certainly has its place for individual shots, if not entire movies.
Some 3D LF films require maximum image quality for exterior beauty shots. Destination films typically use low shooting ratios, and the performance consists of getting the camera movement correct. Prime examples are aerials from airplanes or helicopters flying through a nature scene. Several LF visionaries are improving aerial mounts and gyro stabilization for cameras.
In the 3D world, Ernie McNabb of Kinomax is improving the performance, size, and weight of the only gyroscopically stabilized and remotely steered 3D 15/70 camera system. For 2D films, stabilized rigs such as SpaceCam by Ron Goodman and the GyroPro from John Borden at Peace River Studios give spectacularly smooth and graceful shots.
Ultimately footage shot with these systems might be converted to 3D with the digital processes mentioned above, or the devices may be adapted to the 3D30 or other smaller film or digital cameras.
Jerky, strobing aerials or bumpy dolly moves are no longer acceptable in this medium. Shaky 3D shots seem to be even more annoying than shaky 2D shots. Steadicam is one way to achieve stable, smooth 3D shots with moving cameras. In 2000, Imax, James Cameron, and Stephen Low tried the first 3D HDCAM shots with a Steadicam.
Pioneering filmmakers who believe in “technology push” more than “market pull” are shaping the future of 3D today. Stephen Low has said, “We are, quite simply, a 3D animal, and I believe that is instinctively how we wish to view the world. I have to laugh when people call 3D a gimmick. It’s flat screen cinema that’s the gimmick!”
His father, Colin Low, on seeing Tiger Child, the first IMAX film, reportedly said, “Its flatness is overwhelming…such a big screen needs to be 3D for sure.” We can all be thankful that Colin acted on his vision and developed 3D LF in the 1980s. Stunning examples of superb 3D CGI like Ben Stassen’s SOS Planet underwater animation shown at LFCA show how visually immersive 3D film can be.
Marty Mueller says that what is needed for the future of 3D LF is “Bravery, imagination, a little craziness even.” In Stephen Low’s words, “We are at the very beginning of this revolution. There is hope for 3D just around the corner!”
Gord Harris is a 3D and LF technology consultant with 25 years experience with 15/70 film and digital technology, and has worked on numerous 3D cameras and films. Visit www.go-rd.com or email gord@go-rd.com.
Originally published in LF Examiner December 2002 Vol. 6 No. 2. Used by permission. © Cinergetics, LLC, Columbia, MD, USA www.cinergetics.com
Please click here to read more articles.
|
home | principles | projects | pictures | papers | people | practice | paraphernalia | printed summary
|