|
 |
|
|
 |
 |
|
|
|
|
The focus of the DI Subcommittee, co-chaired by Joshua Pines of Technicolor Digital Intermediates and Lou Levinson of Post Logic, has been to implement the ASC’s Color Decision List (CDL), which will allow cinematographers to assign looks to images that will carry through the editing and other post phases. Working with multiple color-corrector vendors and post facilities, the subcommittee has successfully tested the CDL methodology and continues to work on making it as interchangeable as possible. Pines: “We are in absolute agreement over what underlies the ASC CDL — what the CDL algorithms are, what the numbers are, how they manipulate images — and that is the first step. We had several different vendors and facilities produce the same color-correction manipulations on test images, based on the CDL’s nine numbers. Not every vendor has done this, but the ones that jumped in and gave it a shot succeeded with flying colors. Based on recent discussions, we understand the ones that didn’t join in are waiting to see if there really is a standard. No one has said they wouldn’t adopt it. The question is more about file format. If you’re shooting film and making a print, you can call your lights — ‘Make me a print at 34-27-22’ — and you write those numbers down on your camera report. We’re looking for the equivalent of that. Do we write down nine numbers on a camera report? That’s prone to error. Nine numbers are more than three. Some of them are floating point numbers; some of them have minus signs. It’s hard to write -23.7 and ensure that number gets entered at the other end. “This is the next step: once we have the color correction defined, how do we attach it to a shot? If you attach it to a shot, how do you talk about it so you can reuse that correction on a different shot? There have been a few suggestions floated about file formats. The most popular is XML. The second-most popular is somehow modifying a standard EDL to add a comment field. What’s being debated right now is transport mechanisms; we might end up with multiple transport standards, with some simple ways of translating from one to the other. We want to make this easy for the vendors, but if it starts getting into the internal architecture of different color-correctors, we don’t want to dictate or specify that. “This is just the initial color correction — not, for example, five levels of secondaries and six Power Windows and film-print emulations. This is only for the very basic primary color correction. We’re not infringing on any color corrector’s secrets about how they do tracking or any of the other stuff. The idea is to restore to the cinematographer something he has lost: the ability to assign the overall look very early at the dailies stage. The cinematographer will imprint the first color correction for the dailies that the editors, directors and producers will see all through the editing process. “There is a correlation between one of the CDL parameters and printer lights, the offset parameter. Offset parameter is equivalent to printer-light manipulation. However, many cinematographers wonder, ‘Why are you limiting me to printer lights? I’m on digital now. I want to be able to do gains and gammas, et cetera.’ So we’ve also allowed that. If you limit yourself to using the offset parameter, then you have a direct 1:1 correspondence to printer lights. Some facilities have actually implemented that correspondence for dailies. The concept of digital printer lights is an implicit subset of the CDL. However, once you start using the other things the CDL can do, you get into a world that cannot be characterized strictly by the equivalent of printer lights at the lab. The CDL is more than printer lights, but you can restrict it so that it’s only printer lights. “I think where this is going to come into its own, and we’re beginning to see it, is with the inevitable digital acquisition using digital cameras that capture an extended range. All cinematographers want some sort of look management on set. The ASC CDL is trying to come up with a generic, basic color-correction approach that can establish that look and be carried through the workflow.” Levinson: “[As for the future], the best we can do is recommend good practices and appraise emerging technologies. What we see emerging is an all-electronic pipeline, from capture to distribution. Evidence of this can be seen in the ongoing growth of DI; the Panavision Genesis, Dalsa Origin, Arri D-20 and Thomson Viper cameras; and the growing penetration of digital cinema. How long this takes and what standards need to be put in place to facilitate it are open questions, as are the questions of long-term archiving of the creative author’s intent at the highest possible quality levels. This is not to say there aren’t battles to be fought over whose technology gets used and whose business models work, who controls content at what stages, how content can be disseminated and protected simultaneously, and so on. Because of how fast things change, the standards and recommended practices will always lag behind what happens in the field.” Metadata and Digital Camera Subcommittees The Digital Camera Subcommittee, chaired by David Stump, ASC, is working toward the ASC Camera Assessment Series, a repeatable series of typical feature-film-style shots to display the performance characteristics of the various digital-cinema cameras, thus enabling filmmakers and others to make more informed choices. Metadata, which can contain limitless production information, will be an integral part of the digital processes, and the Metadata Subcommittee, co-chaired by Stump and Ana Benitez of Thomson, has been working in an open forum to find ways to enable metadata’s use in applications. Stump: “The Metadata and Digital Camera subcommittees are deeply intertwined. In 2003, I put together the Metadata Summit at NAB, a private meeting between the Technology Committee and people from the manufacturing community and community at large. It was a discussion to advance the topic of metadata as a serious subject for cinematography and our industry. Our discussion was pretty fruitful, and I think it raised a few people’s consciousness on the subject. What it really did was put metadata on the tip of everybody’s tongue. “The golden rule of metadata is: add to it, but do not delete from it. Evidence that the rule is sinking in is showing up in the committee. At the last committee meeting, a hardware/software company called Monster showed up. The company’s goal is to create precisely the on-set computer I called for in my Metadata Summit white paper three years ago, “Film Production in the Future: Transparent Metadata in an Automated Workplace.” Monster has invented a computer that will collect this data on set, and it is working toward inventing tools that will allow people on set to contribute to the database. “The most difficult position to interface with seems to be the script supervisor. That job is pivotal to the notion of collecting metadata but is also the most mechanically difficult. That’s the nut we have to crack. This will mean big changes for script supervisors, but not bad changes. It will take a very well-conceived and well-executed system to win over script supervisors. The language and mechanism by which they do their work is difficult to put on a computer — the way they line a script, the way they keep notes. When we have a breakthrough in that area, it will revolutionize the whole process of collecting metadata on set. “Nancy Wilkerson at Disney Studios has a project called the Black Diamond Project that is specifically trying to crack that problem with hardware and software. Monster is trying to build those devices. Everyone knows this is an eventuality; they’re just waiting to herd it in with a common language and common understanding of how it needs to work. When script supervisors realize that it will strengthen their position on set, I think they’ll embrace it, and once they do, we’ll cross that tipping point.
|
|
| |
<< previous || next >>
|
|
|
 |
|
|
|
|
