One of the big challenges created by digital cinema technology can be summed up in a single word: Frozen. Disney’s wildly successful 3D animated film captured two Academy Awards – Best Animated Film and Best Song – and was an international box office success. And therein lies the problem. For hit movies like Frozen all those various markets demand literally thousands of different and unique versions and producing them under demanding deadline pressure is a huge undertaking. “It seems unbelievable, but for one title that is a scope show (2.39) aspect ratio, we usually have to create a 1.78 panscan and a 1.33 panscan,” said Annie Chang vice president, post-production technologies, The Walt Disney Studios. “When you start to factor in 42 different languages for each aspect ratio and then add in multiple resolutions/playback standards (1080p, 720p, NTSC, PAL) and then add in the different edits/versions and the various delivery formats that downstream distributors (airlines, television, cable, satellite, electronic sell through, etc) ask for, you exponentially end up with tens of thousands of files. Just for one title.”
Chang made that point when she spoke at SMPTE’s Technology Summit on Cinema at the National Association of Broadcasters convention in April. The figure “tens of thousands” astonished me and I asked her to confirm the number, which she did when she sent me the email quoted above. To address this situation, SMPTE is currently working to develop an Interoperable Master Framework (IMF) standard. A 2013 white paper written by AmberFin explains how the IMF standard would work:
“In essence, the Interoperable Master Format (IMF) is an evolution of the DCP (Digital Cinema Package) architecture, providing a complete file interchange unit to the distribution channel. DCP is really about finished content for theatrical presentation. IMF is different. As content makers look to generate the maximum revenue from their products, via the maximum number of bums on seats in as many markets as possible, this means creating multiple tailored versions of the same piece of content for each of those audiences.”
“A film like Toy Story 3 for instance would involve thousands of different main and promotional versions for each individual market, as well as airline and television edits, etc. With IMF, you wouldn’t need to create a thousand plus copies of the content; what the standard does is separate the content into various ingredients or components (namely, AS02 MXF media files), a number of ‘recipes’ (Composition Play Lists) and a selection of instructions (or Output Program Lists) appropriate for each of those audiences. IMF is designed to take the right mix of ingredients, the right recipe and a tailored set of instructions to create a dedicated version for each market, without having to duplicate files.”
“We have always assumed that IMF would eventually support some type of encryption model,” says Disney’s Chang. “In fact, in the early days of development, many of us content owners had just assumed that we would adopt what was used in digital cinema. However, once we nailed down a few of the core specifications needed for IMF at SMPTE, we discussed in the Working Group whether encryption was a priority or whether we should work on other core functionality like Output Profile Lists and interchange instead.”
“The content owners agreed that IMF was and is different than DCPs,” Chang continues. “We are distributing IMFs to trusted vendors downstream who have contractual obligations with us. It is not the same as putting a DCP out into the world with tens of thousands of theaters – we’re just distributing IMF to a few companies downstream. We wouldn’t encrypt an HDCam-SR (if you could), so we also see no need right now to encrypt an IMF. Of course, this could change in the future, so we have left the door open for discussions around encryption; however, right now, we are focused with getting our basic requirements of achieving a truly file-based workflow (and not just a tape-replacement file) with IMF.”
Mark Gray, CEO of Archimedia Technology says, “The whole concept of digital cinema is still very much in its infancy and evolving. Things that the studios worried about a year ago are quite different today. Workflows are improving with new ways of accessing digital cinema files such as the Archimedia Master Player. Also, there are new file formats and packaging structures that will help in the production and distribution of digital movies. We particularly like IMF, which allows a packaging of one essence file with all elements of a movie. An instruction set is then added to the file, which allows an intelligent sorting of the appropriate visual and aural elements for distribution. Today movie studios literally make thousands of variations of a movie. IMF potentially will drastically reduce cost and increase quality. Archimedia is happy to be one of the leading companies developing IMF products.”
If digital cinema technology, and DCPs in particular, has created some serious challenges for the major Hollywood studios, the same is not the case for the broader motion picture production business. In an ever-expanding number of applications DCPs have evolved into an increasingly popular method for exchanging files in virtually all aspects of the broader motion picture industry including event cinema, independent film, film festivals, broadcast, corporate video and education. “The platform is ubiquitous now,” says Stuart Bowling, director of market development, Dolby Laboratories.
There is also a growing understanding that encryption and some other restrictive aspects of the DCP spec aren’t a critical concern in those markets. In fact, encryption protocols have prevented many independent filmmakers from showing their movies at festivals often because the keys have simply expired.
For the Hollywood studios the security of their content was – and understandably remains – one of their greatest concerns. The DCI mandated tight encryption protocols that have, to a degree, helped stem the problem of movie theft. Over time, though, outside of the major Hollywood studios it became obvious that they keys created a host of real world problems, many of which could be traced to human error of some kind. As the industry has grown accustomed to digital technology, the need for a more flexible approach to encryption keys has developed.
Mark Rupp, president and COO of SpectiCast, a Philadelphia alternative content company says, “My background is in technology and when I saw that DCI systems were going to be a reality for traditional cinemas, we immediately added this option to our list of delivery platforms and now it makes up the majority of our bookings. While we are somewhat different than many of our competitors that only focus on cinemas, we are also the only U.S. based event cinema distributor offering content on an international basis. In the U.S., we offer our programs to a wide variety of exhibition venues such as performing arts centers, museums, concert venues, and even retirement communities. For those venues we continue to supply service over our own Digital Theatre Network or Blu ray. For cinemas we provide DCPs and in fact have the capability of encoding, duplicating and distribution in house including satellite delivery of DCPs to many of our European venues. Over time, with technologies like DCDC and SmartJog, delivery of DCPs over the air will become more relevant.”
Understanding that encryption is not always a positive, and that it can even be a headache explains Adobe’s approach to digital cinema packages. Just before this year’s NAB Adobe announced that it would be working with digital cinema pioneer QuVis to integrate the QuVis Wraptor digital cinema package encoder into Adobe Creative Cloud.
“DCPs offer a unique solution for delivering and storing ultra-high quality content, QuVis CTO Kenbe Goertzen said at the time. “We see this integration with Adobe as a significant step toward bringing DCPs beyond theatrical movie distribution and into the mainstream. We believe that over the next 12 months DCPs will be used more and more widely for dailies, independent distribution, archive, and other applications.”
I spoke with Adobe senior product manager Patrick Palmer right after NAB to learn more about the new partnership. He said Adobe had discovered that the interest in DCPs was much higher and broader than they had expected and included independent filmmakers, schools, broadcasters and corporate video producers. These customers also had some shared complaints about DCPs, namely that they aren’t always compatible with every application and, most commonly, that they are too hard to make. Palmer says that although Adobe did not have a precise number for how many potential DCP customers they might have, the number was huge. Palmer praised the engineers at QuVis for delivering a product under impossible deadline pressure. “QuVis is really interested in meeting our paramount concern, which is simplicity,” says Palmer. “Believe me, that wasn’t easy.”
He neglected to add that the other mandate that Adobe gave QuVis: the product had to be available to customers for free.
Keith Doughty, QuVis vice president, sales fostered the new Adobe-QuVis partnership and despite all the hurdles his engineers faced he says, “Adobe is a great company to work with.”
When Doughty attended the SMPTE Fall conference in Los Angeles last year someone suggested that Adobe might be interested in partnering with a DCP encoder company. He called Patrick Palmer at Adobe and said it was a case of “right place, right time” because they were interested. The two men met last December and Palmer told Doughty that Adobe had two requests: they wanted to announce the partnership in April at NAB and they wanted “a functional DCP with Media Encoder at no additional charge.”
“For us that was pretty short notice,” says Doughty, but he knew it was a great opportunity so his answer was, “Let’s see what we can do.”
The resulting product is 2K only and does not have all the functionality of other DCP encoders on the market. That was by design. “We think there’s a better way of doing it. Encryption can be a drawback,” Doughty says, “so we took encryption out of Wraptor. It’s not going to cost $3,000 to create a DCP.”
QuVis, in addition to working with Adobe, is still in the DCP market on its own. “We believe there are opportunities on the playback side,” Doughty says. “We’re looking at ways of improving our player. We want to drive down the cost of that player.” He believes that every DCP generator will lead to the demand for ten players. To that end QuVis is developing a rental market for its DCP Player v.1 that can be rented for $50 a month. Options in a fully featured Wraptor include 2K and 4K, and multiple frame rates of 24, 25, 30, 48 and 60 frames per second.
At this point there are a growing number of companies, large and small, developing software and hardware systems built around the DCP and IMF concepts. This competitive atmosphere should lead to continued advances in how motion picture files are created, shared, distributed, exhibited and saved.
Kevin Wines, director, cinema products, Doremi praised the work that the members of the Digital Cinema Initiative did in creating a great framework and says it was understood from the start that DCPs were not a finished concept. “We have to give lots of credit to DCI but DCPs are a natural evolution,” he says. “We’ve only begun to tap the potential.” As just one example, Wines said one potential application that’s just getting looked at is to use DCPs for archiving. “It’s a way to practically keep everything forever,” he said.
“Digital cinema hasn’t been realized,” says Dolby’s Stuart Bowling. He calls the current efforts to incorporate IMFs into the DCP framework “part of the evolution” of the digital cinema package. “We’ve only scratched the surface,” Bowling says.