In Part 1 here, we detailed the history that led up to the establishment of the Advanced Technology Committee of the Academy's Science and Technology Council. The goal of the Committee's Interchange Format Project is to create an interchange format that would allow for digital color management for motion picture production/post.

Andy Maltz, director of the Science and Technology Council, wants to dispel one idea that might have been extrapolated from last month's article. "It is likely that there'll be colorimetric encoding, but we also want to make it clear that we're not restricting workflows," he said. "The goal is to enable a minimum 4K wide color gamut workflow, but also to allow people to use the workflow they want."

What Maltz is referring to is the fact that numerous mastering facilities work in printed density spaces (Cineon essentially uses densitometric information). "I didn't want readers to get the sense that we're designing that out," he said.

What the Committee is doing is defining "interface points" (or the conversions going into and out of the digital color space) and "data representation formats" (synonymous with the color encoding scheme). "The image encoding space is the numbers that represent the colors," explains Maltz. "The interface points are the transforms or conversions as the image goes into and out of the image encoding space."

Right now, the Committee is evaluating rendering transforms at those interface points. "The process makes assumptions about creative white point and the ambient light levels and renders it for a class of display devices," said Maltz. "We're focused on collecting a set of reference images that eventually should be made available to the broader community. Besides the math and science, at the end of the day, it has to work on an agreed-upon set of images that have all the agreed-upon characteristics for motion pictures."

In addition to charts, those images would be scenes that were created in a "known and agreed upon" way. "That's what you're anchoring the implementation of the rendering transform to," explains Maltz, who notes that, for example, it would be reasonable to have a set of reference images originated on the commonly used Kodak 5218 negative stock. "If you use the wrong images to test out your rendering transform, you could be building in some things that are mistakes."

But those publicly available reference images are a way off. "It's hard to communicate what we've been working on for two years," said Maltz. "It's a similar problem to communicate the results to the general community and provide them with the tools and reference materials to implement those results. You have to figure out how to roll it out and get rapid and widespread adoption."

The Committee is not a regulatory body, however, but it's been in close contact with SMPTE (Maltz chairs the ad hoc group on digital leader). "Our longer-term goals are to take whatever we've done in committee and bring them to a standards body such as SMPTE."

Therein lies the rub. While the Committee wades through the difficult and laborious work to create best practices that can be standardized, manufacturers continue to sell DI solutions and DI artists continue to color motion pictures. "There is a window of relevance for this sort of work," agreed Maltz, who reports that manufacturers are kept apprised of the Committee's progress. "How long is that window open for? I can't tell you, but we're cognizant of that. At the same time, this sort of work demands that you don't rush something out the door. It's not like the commercial world. But if you take too long, the decisions will be made by the marketplace."