If we were to define in simple terms the current role of the SMPTE 21DC committee, it is to oversee and improve all things related to digital cinema distribution. 21DC oversees matters related to encryption, including the selection and details of encryption algorithm, the Key Delivery Message (KDM), and, of course, the DCP. It also oversees specifications that affect post production, such as standards for interoperability of intermediate files, including white point.
But 21DC’s primary responsibility is the SMPTE DCP format. SMPTE DCP may be the only format that is tested by DCI compliance testing, but it is a format that has yet to be successfully introduced in real movie distribution. Without regard for its lack of use, SMPTE 21DC continues to develop the SMPTE DCP format, which only further complicates its introduction into distribution.
21DC has revised the core specification of SMPTE DCP, SMPTE Standard 429-2, twice since its initial release. A notable addition to the original specification is the addition of Multi-Channel Audio (MCA) labeling, which does not accommodate 3D or “Immersive” audio, but provides a more flexible manner of wrapping audio channels in the DCP. Notably, MCA also requires more flexibility in audio routing, which should raise a flag. Further, there has recently been talk of revising SMPTE Standard 429-7, the specification for Composition Playlist (CPL). This is a lot of work, a lot of investment by the manufacturers that support the standardization process, and a lot of upgrades that will be pushed on exhibitors, for a distribution scheme that so far has proven to be unsuccessful. Even DCI only supports parts of the specification, and has yet to embrace the full scheme as described by SMPTE. SMPTE DCP, in short, has become an engineering marvel, shunned by every distributor that wants its movie to play.
Undaunted by its lack of success in the field, SMPTE 21DC continues to develop SMPTE DCP. This month, Dolby proposed a new track file into which it can carry its proprietary Atmos audio. It is also asking for a new synchronization scheme to be standardized, that goes beyond the Ethernet-based scheme adopted for closed captions. Even if the standards are realized, it’s easy to predict the outcome: the Interop format will be modified to accommodate them, and SMPTE DCP will remain a novelty.
If there is need to further this point, then one should consider HFR. SMPTE 21DC this month reviewed plans to update specifications to incorporate higher frame rates. Interop DCP, out of necessity for the HFR 3D release of Hobbit, has already been updated.
The biggest challenge to SMPTE DCP is the installed base of systems that offer limited support. Stated differently, there is a substantial base of systems in the field that will never support full SMPTE DCP capability. As the SMPTE DCP specification continues to evolve, more and more systems fall into the camp of “no support.” Unless distributors were to proclaim that they’d pay exhibitors to upgrade, this scenario is unlikely to change. In the real world, exhibitors have no incentive to upgrade systems, and distributors have no incentive to risk switching from Interop DCP distributions to SMPTE DCP. The goal of both parties is to play the movie, and that goal is met with the specifications in use today.
Most importantly, where it was once believed that older systems would be phased out, leaving behind only SMPTE DCP compliant systems, that belief is now challenged by the continued evolution of the SMPTE DCP specification. There will always be a significant base of installations that cannot support the latest and greatest from SMPTE. And there is huge risk that the most valued components of SMPTE DCP will continue to find their way into a revised Interop specification, as that path offers the lowest risk to distribution and exhibition. The lowest common denominator will always win.
To get beyond the lowest common denominator problem, better content management practices are needed, where databases are built to contain accurate information as to the capabilities of every installation, allowing the distributor to accurately target more capable formats to more capable cinemas. If anyone should know this well, it would be Warner, having experienced first-hand the intense content management problem introduced with the HFR 3D release of Hobbit. But no voices of this sort are emerging.
The result is that SMPTE DCP continues to evolve blindly, without any reality checks applied. A point to consider is that there are no rules in SMPTE to preclude the standardization of alternative formats. Perhaps one day someone will suggest standardization of Interop DCP. Then the reality check will begin.