Setting Standards for High Frame Rate Digital Cinema PART 2
COW Library : Stereoscopic 3D : Michael Karagosian : Setting Standards for High Frame Rate Digital Cinema PART 2
I was asked by SMPTE to co-chair the HFR Study Group, presumably because of my experience in chairing committees and my work in exhibition, and certainly not because of some fabled ability to discern image artifacts. I jumped in because I knew I'd have the chance to work closely with two talented cinematographers, Kommer Kleijn, S.B.C., and Dave Stump, A.S.C. HFR only makes sense to me if driven by the creative community, and our combined skill sets would drive the study effort in the right way.
Dolby Split Surround encoder/decoder for 70mm film was introduced with Apocalypse Now.
I am quite familiar with the exhibitors' point of view, having consulted to NATO for 11 years, and having spent much of my career developing and marketing, as well as helping others develop and market, new products and services to the exhibition market. The introduction of digital cinema was challenging. It was an unbalanced proposition, and up until 3D, it was only a me-too experience. Distributors would save money in fulfillment, but exhibitors would spend more on equipment. Technology companies were attempting to build momentum through the excitement of this new technology, but the truth is that the reason we go to the cinema is to see a movie. We don't go to the cinema to see a projector.
Fortunately, it happens that a single digital projector can also show high quality 3D images. What really clinched sales for digital projection is that the creative community loves 3D, and exhibitors are able to charge more for the experience. It was this unique combination of forces that drove the adoption of digital projection in cinemas.
The big challenge today for technology companies is the massive downturn in sales that is destined to take place the last half of this decade as the digital installation boom ends. To counteract, the natural instinct for these companies is to prey upon the human desire to have the latest and greatest. For this reason, you see manufacturers pushing HFR technology, new sound capability, higher brightness, lower cost devices, and on it goes. It's not possible, however, to achieve the same overwhelming welcome by exhibitors that 3D enjoyed without the means to monetize new investment, an aspect that is sorely lacking with newer technologies.
The push for HFR has more twists to it. Filmmakers understand the new capabilities of their cameras, and maybe they understand the production workflow bottlenecks that new digital cameras impose, but few if any go as far as to understand the impact that they will have in exhibition.
The drive for HFR with Hobbit has also been ahead of the game, causing many manufacturers to drop development projects that would have been more profitable for them. As the time for movie release approaches, real problems in the field led to a legitimate concern on Warner's part that a wide-scale HFR release will cause dark screens, or just as bad, improperly configured presentations. That led Warner to limit the number of HFR releases to 400. For all the efforts made, 400 screens is not a very rewarding market for either manufacturers or exhibitors.
More attention needs to be paid to the creative drive for HFR. The press (obviously not Creative COW) likes to say that all movies will be produced in HFR. Just like the press used to say that all movies would be 3D. But like 3D, HFR is just another tool for story telling. Doug Trumbull, ever the world champion for HFR, having been the pioneer behind SHOWSCAN in the 80's, talks about HFR as a tool to drive emotion and intensity, as story-telling tools. Even Doug doesn't talk about HFR as the right tool for every story. This mixed world of frame rates imposes an interesting set of challenges on exhibition systems. And the creative exploration of higher frame rates is only just beginning. There is more to learn, and it is this uncertainty that poses a problem for manufacturers, who have to tow a line. If they put too much technology into their products, they'll drive up price with a negative impact on sales. If they put in too little, the new features won't be useful to filmmakers, and of no value to their customers, the exhibitors.
Filmmakers understand the new capabilities of their cameras, but a very present problem is the impact in exhibition. HFR poses legitimate concerns such as improperly configured presentations or dark screens. This potential problem led Warner Bros. to limit HRF releases to just 400 screens.
Ready or not, a stake has been placed in the sand. Image content is compressed using JPEG2000, at a visually lossless rate. So that filmmakers have a concrete ceiling to work with, and so manufacturers could build competitive and interoperable products, DCI (Digital Cinema Initiatives, a consortium of the 6 major studios) specified that products must support a minimum of 250Mb/s for compressed image. But that spec was not to exceed 4K at 24fps, and even then, the cinematographer community didn't feel that the 250Mb/s number was enough. When we get into higher frame rates than those specified by DCI, we step into new territory. Today, the latest generation of products supports up to 500Mb/s of compressed image data. But the question is still there as to whether this is enough.
For The Hobbit, 500Mb/s is considered sufficient, but only just. The HFR version of Hobbit will be released at 48fps per eye, or 96fps if you're a projector and simply counting frames. Doug Trumbull and others have determined that the optimum impact on our senses occurs at 60fps for 2D, or 60 fps per eye (120 fps) for 3D. To get there and preserve quality, we will need new development. That development could be in the form of smarter encoders that better manage bit allocation among image objects during compression, or the introduction of decoders in the field faster than 500Mb/s, or both.
My guess is that we have an installed base today of around 10-15K 500Mb/s decoders, if all of the upgrades were to be implemented, out of over 80,000 digital projectors in cinemas around the world at this time (Oct 2012). Both counts will grow with upgrades and new equipment sales as the 40,000 film screens that remain are converted. Obviously, if you get the drift of where this is going, we're going to have a large base of 500 Mb/s HFR-capable equipment, and it would make a lot of people happy if we could simply improve encoders, rather than push for another generation of decoders in the field.
Filmmakers have a concrete ceiling to work with compressing using JPEG2000, a visually lossless compression. By stepping into higher frame rates than DCI's standard minimum support of 250Mb/s for compressed image specs of 4K at 24fps, we're looking at requirements for smarter encoders, field decoders faster than 500Mb/s, or both.
However, we first need to better understand what we have today, and what cinematographers need. To get there, we need HFR test material. You would think that we would have tons of material to work with in Hollywood, but all of the good stuff has rights attached to it. We need a wide variety of carefully designed, real-world shots to stress JPEG2000 encoding at higher frame rates. I emphasize "real-world" as we could use pathological test patterns for this purpose, but we would get pathological results, pointing us to system requirements that may be way beyond what we need for real-world movie making.
The availability of practical HFR test material, in a big way, will help us solve two problems at once. The first problem is we need to comprehend the restrictions imposed on bit rate, frame rate, and quality level in our 500Mb/s world. The second is that we need to empower manufacturers to build better encoder and decoder products. By better I mean smarter encoders, and higher bit rate decoders. Open, real-world HFR test material will enable manufacturers to be more innovative and sure-footed in their product development.
The HFR Study Group is very much a worldwide, collaborative effort, with participating members from several continents. The group has two major tasks. The first is the cataloging of system capability, existing system issues, and areas where more testing is needed. A summary of that work to date was published in the SMPTE Journal's September 2012 issue in the 21DC report. The second is an evaluation of the limitations of current equipment from a quality perspective, and the limits that the industry should target. To get to the second step requires the test material I spoke about. A fair chunk of my time is spent leading our Test Team group, centered on the production planning for the shoot of HFR test material, and the steps needed to evaluate it. My hope is that by this time next year we can talk about the completion of this work, and the outcomes that we will have observed.
One of the original chairs of SMPTE's Digital Cinema standards committee, Karagosian has led several DC subgroups, including the original Packaging group and the former Exhibition group. With David Stump, A.S.C. and Kommer Kleijn, S.B.C., he now co-chairs SMPTE's High Frame Rate Study Group. He spoke to Creative COW about his work on that committee as well as his thoughts on 3D, digital cinema and HFR.
Thanks also to Debra Kaufman for coordination and additional editing on this piece.
Special consideration to Warner Bros. for images from "THE HOBBIT: AN UNEXPECTED JOURNEY".
MARTIN FREEMAN as Bilbo Baggins in New Line Cinema's and MGM's fantasy adventure "THE HOBBIT: AN UNEXPECTED JOURNEY," a Warner Bros. Pictures release. Photo credit: James Fisher. © 2012 Warner Bros. Entertainment Inc. and Metro-Goldwyn-Mayer Pictures Inc.
HUGO WEAVING as Elrond in New Line Cinema's and MGM's fantasy adventure "THE HOBBIT: AN UNEXPECTED JOURNEY," a Warner Bros. Pictures release. Photo credit: Todd Eyre. ©2012 Warner Bros. Entertainment Inc. and Metro-Goldwyn-Mayer Pictures Inc.
Peter Jackson's King Kong from Universal Pictures. ©2005.
Francis Ford Coppola's Apocalypse Now from United Artists. ©1979.