If there was a rule to life, it would be that no one can escape change. Today, server companies that once thought of themselves as fortresses built of stone are now looking at foundations built on less stable soil. Projector companies, once partners in sales efforts, are now out to eat their lunch.
It has long been a tenant of this publication that the media block was destined to become a commodity. Server companies, driven by engineers that are absorbed with the intricacies of their products, could not see how this would happen. But the decision makers in their customer base, who are only interested in playing movies, know little about intricacies, and ultimately look upon equipment as a line item on a sheet of paper. If the projector company offers an internal media block and server, then that line item can be checked off.
Sony, enamored with vertical integration, never opened the door to 3rd party server companies. TI, however, strategically chose the opposite approach, creating a substantial market for 3rd party server manufacturers. So much so, that the digital cinema landscape is littered with server companies that didn’t make it. Those that did make it have a lot to be proud of. But the picture of success is changing as DLP projector companies develop their own internal media blocks that will sell as a commodity component of their projectors. Intopix, producer of cinema-specific configuration software for off-the-shelf field-programmable gate arrays (FPGAs), makes this easy to do.
Each company, both server and projector, have shared remarkably different strategies for how they are addressing the future. Taking projector companies first, Christie views its strength as a hardware company, and so developed its own media block to be used with off-the-shelf servers. Barco is thinking differently, and sees its future in software. This is what drove its purchase earlier this year of XDC’s software assets. Combining strengths in both hardware and software could give it a more “Apple” feel, for want of an analogy. But Barco has not demonstrated its own media block as yet, instead demonstrating OEM media blocks from Mikrom and USL, which it can re-brand. Its long term thinking about Barco-built media blocks has yet to be learned. NEC has the potential to be strong in both areas. But potential is a big word, and while NEC has developed its own media block and software, it hasn’t impressed US exhibitors with its abilities. This may, in large part, have less to do with the quality of its engineering, and more to do with the belief, or lack of it, that NEC will be in digital cinema for the long term.
Server companies have their own strategies, none of which are convincing. Doremi is the only stand-alone server company that builds its own media block, giving it the bargaining chip of licensing both design and brand, that other server companies do not have. Dolby, with its on-again/off-again commitment to cinema, gave up building media blocks a few years ago, courting Mikrom and USL for its in-projector media block strategy. It appears, though, that even this decision is subject to change, and it now recognizes that without a media block, it has nothing to sell projector companies. Dolby, however, very likely has a grander, if not riskier, strategy at play. Expect Dolby to couple its new audio recording method with its media block and server, to give its entire cinema product line the integrated, proprietary nature that Dolby seeks. More on Dolby’s audio plans in the audio chapter of this report. GDC has always been more focused on software than on hardware. It now has an exclusive arrangement with video processing expert AJA to produce its media blocks. Not surprisingly, GDC’s former head of R&D now works for AJA. GDC’s clout will be built on its connections in Asia, introducing cooperative projector companies to sales. Qube also utilizes the Mikrom media block, but has more presence in eCinema than in digital cinema. If the Indian market should upgrade, Qube has similar opportunities there, just as GDC has opportunities throughout the rest of Asia.
Projector companies are positioned to be in full command of the system sale, and it’s difficult to think that any effort from a server company will unseat that. Where things are less clear is in the area of software. To an exhibitor, it’s software that manages the booking, that performs the scheduling, and handles reconciliation with the distributors. It’s software that will bring automation to the receiving of content, and to security key management. All of which makes the software partner a critical consideration. Among server manufacturers, no one stands out as a potential winner in the cinema software market. Among projector companies, only Barco has positioned itself to tackle this area. There could be surprises in store, as events unfold.
One of life’s little twists that exhibitors have been dealt is that stand-alone servers are limited in the capabilities that future blockbuster movies will take advantage of. While stand-alone servers populate the majority of digital cinemas, the limited bandwidth of their dual-link SD-HDI connections are incapable of higher frame rates. Only in-projector media blocks can perform this feat, and it could turn out that even these have hardware limitations. No formal exploration has taken place to determine the compressed image bit rate required to achieve visually lossless images at higher frame rates. A simple numerical calculation says that if 250Mb/sec is good for 24fps per eye, then 625Mb/sec should achieve similar results for 60fps per eye. But few, if any, in-projector media blocks can achieve this.
Any attempt to revisit compressed image bit rates will cause cinematographers and other golden eyes to revisit the issue of visually lossless. It’s an old complaint that 250Mb/sec is not enough for 4K. This causes some to think that 800Mb/sec may be needed for 60fps per eye 3-D. SMPTE’s HFR Study Group is not authorized to conduct its own high frame rate tests, but with two cinematographers as co-chairs with your author, it’s reasonable to expect that no stone will be unturned. SMPTE may not be able to conduct its own tests, but it can recommend how to do it, and collect input from those who have done it. This won’t happen overnight. But some effort will take place to quantify the bit rate that new high frame rate products should achieve. (This is unlikely to become the basis of a new DCI spec, for reasons covered in previous reports.) In the short term, the SMPTE HFR Study Group will learn what’s available in the marketplace, in the interest of helping studios manage the near-term release of high frame rate movies, such as Hobbit.
Support for high frame rates, unfortunately, will not present a wedge that server and media block companies can use to push aside their new competition. Most all products today use the Intopix software configured FPGAs, where a change to higher speed parts allows the processing of higher frame rates. Server and media block companies now find themselves in the position of having to improve their products, while figuring out their futures.