SMPTE is getting on the higher frame rate bandwagon. Kommer Klein, a European-based cinematographer, several months back initiated a new work statement for high frame rate 3-D within the SMPTE 21DC Technology Committee for Digital Cinema. The proposal received a cool reception at first, and while improvements to the scope were suggested, no urgency was indicated by the larger membership.
But that was eventually to change. Early this month, Wendy Aylsworth, of Warner Bros., posted a significant revision of the work statement, followed by supportive comments from Fox and Paramount. (When one studio speaks and a few others chime in, it probably means there was a DCI meeting the day before.)
In the course of discussion, some interesting facts emerged. The modified work statement will lead to a study effort, which should include visual testing of different decoder bit rates. (Note that decoder bit rates are independent of frame rate. A high frame rate coupled with a low bit rate will result in low picture quality.) Your author attempted to include 60 fps 2-D in the tests, to complete the study of appropriate bit rates for non-DCI frame rates, and was told by Mr. Klein that the 60 fps was already decided. This was news not only to your author, but to Fox and others who work closed with the American Society of Cinematographers.
It turns out that the Joint Photographic Experts Group (JPEG) very recently extended the JPEG 2000 standard to include 60 fps. Conveniently, JPEG chose 250 Mb/s as the determinant for the JPEG 2000 codestream. This was set without consultation with the cinematographer community. To complicate matters, Mr. Klein claims that SMPTE established 250 Mb/s as the bit rate for 60 fps by interference of SMPTE 428-11 Additional Frame Rates. In this, Mr. Klein is wrong. SMPTE has not issued a standard specifying decoder bit rates in digital cinema, and the JPEG work discussed did not exist when 428-11 was issued, so no such interference is possible. The sad outcome is that, between JPEG and Mr. Klein’s actions, the industry will not get a chance to have input on factors that determine image quality at 60 fps.
The sudden interest by major studios in having SMPTE undertake a study of bit rate for higher frame rates is noteworthy. A possible agenda is to have SMPTE standardize minimum decoder bit rates. It would make sense as DCI is not in position to change its spec. This would be a good step if the specification is based on visual testing that engaged the cinematographer community.
SMPTE 21DC is also moving down the path of producing a revised on-screen subtitle standard that allows for control of the z-axis in 3-D subtitles and captions. The parameters for communicating z-axis information have been well-studied and should not be controversial. The manner for introducing the 3-D subtitle standard to the field, however, could be. Changing a standard across over 50,000 screens is bound to be problematic. There is question as to whether SMPTE 428-7, the current 2-D on-screen subtitle standard, is kept as-is, and later deprecated after a new 3-D standard is introduced, or to simply update 428-7 to the new 3-D standard, if backwards compatibility can be maintained. (For those familiar with XML, differences in schema and namespace are likely to interfere with backwards compatibility.) A constrained version of 428-7 is also used when packaging closed caption text, so backwards compatibility with that application also needs to be considered.
