Cameron employs CFI Laboratories' unique remote color-timing system.
While feeling inventive one day, Consolidated Film Industries' head color timer, Dan Muscarella, imagined a procedure that would allow him to utilize the Los Angeles-based company's INPS (Integrated Negative Preparation System) computer timing system while working outside the confines of the laboratory. As a result, the timing process on Titanic could be concluded during sessions with director James Cameron at Skywalker Ranch, located in San Raphael just north of San Francisco. The filmmaker worked very closely with Muscarella while overseeing Titanic's sound mix at the Ranch's famous audio facility.
The INPS program, based on existing software more commonly used in Europe, collects a footage and frame-count address during print projection, and can be input by the user with coinciding RGB timing-light settings, effects notations, and scene descriptions. A finalized INPS file is output as a paper tape, which delivers the instructions to CFI's printers. (In the future, after the printer controls are integrated with the lab's computer system, the digital file itself will run the machine.) The system also allows the user to link shots together, attaching the timing information to any cutback from the same take. Any future timing adjustment will then be automatically updated through the chain.
CFI technical director Richard Smith adopted Muscarella's remote-timing idea and began to plot a procedure that would make it a reality. First, an IBM-compatible laptop computer and desktop docking station was acquired. The docking station was necessary to accommodate the full-size I/O card required by the INPS program. Next, the laptop was installed with INPS, Microsoft Word, and Quicklink II. In addition, a film FCC (frame count cue) interface needed to be installed on any equipment on which the film was to viewed. The FCC would communicate with the INPS program and lock Muscarella's timing notations in sync with the appropriate footage.
For timing purposes, films are normally viewed in a small screening room. However, Cameron prefers to time his films on a three-picture-head KEM table, with the aid of Muscarella and cinematographer Russell Carpenter. "We have the edited work picture on one head, the new print on the center head, and the previous print on the third," the director explains. "We use a foot-lambert meter to true-up all the heads in terms of intensity, and a Kelvin meter to true the color although Dan sometimes just does those things by eye, which I think is actually better. He uses very fine grades of optical-printer filters to true the color between the various bulbs. From there, we can go though the picture on a frame-by-frame basis, making notations as we go."
The KEM could be easily attached to the INPS FCC counter by way of a custom linkage of Smith's design. With this hardware in place, he planned the physical process that would enable Muscarella and the CFI lab crew to stay in sync despite the nearly 400 miles between them. First, each reel of original cut negative, whether it was an A/B or an A/B/C roll, was programmed into CFI's INPS system. Scene by scene, information regarding FCC, RGB lights, effects, and scene descriptions were input. Once this was completed, CFI staffers struck a print of each reel and created respective diskettes containing the coinciding INPS data.
This package was flown to Muscarella and Cameron at Skywalker for their evaluation, with the print on the KEM and the corresponding INPS file in their laptop. After their color corrections were completed, the updated INPS data was sent back to CFI from their laptop via a fiber-optic telephone line. A new print was then struck at CFI using the updated INPS data, and the cycle would be repeated. Special requests could also be faxed to CFI directly from Muscarella's computer. (As CFI's computer system is refined, the need for shipping diskettes with the prints will be eliminated and the remote laptop will directly access CFI's main server.)
Pleased with the KEM/INPS system, Cameron states, "This method is more accurate than any system that the labs provide. Most timing takes place as you sit down in a screening room. They run the print and you point at the screen saying, 'That shot, no, the one before it, the one where she came into the room' By then you're at the end of the reel, you've missed 10 different things and you have to watch it over and over again. They just sort of wear you down. And if you're doing complex work and assigning one light to a shot that goes from cut to cut, some labs will typically time from a single frame of that shot. Well, that light might have changed two frames later, and it could be completely bogus. All of this means that it will take more run-throughs to get the timing right, which means more wear and tear on your negative. Dan and I feel that on a typical reel of Titanic, one without a lot of effects, we can get to our final light in a hazeltine and two first trials. That's pretty low.
"Dan and I used the INPS system on T2 and True Lies [while working at CFI], but the main improvement here is that instead of running back to the lab with his printer information written on paper, he can just modem the data right out of his laptop. That makes it much more precise. And having the laptop tied directly into the KEM so that the timing information scrolls in time with the film as we watch it works beautifully. It's a much quicker process."