Remote HD Engineering
May 24, 2005 4:27 PM, By Michael Goldman
As its title suggests, much of the new feature film The Cave, directed by Bruce Hunt, takes place in a wet, mysterious, dark cave. In fact, much of the movie was shot in a real cave in Mexico's Yucatan Peninsula, with additional scenes shot in a giant water tank in Romania. Along the way, filmmakers developed a sophisticated HD production workflow to shoot efficiently in both environments.
To sort out the best way to shoot in Mexico and Romania, DP Ross Emery turned to HD digital imaging engineer Nick Theodorakis. According to Theodorakis, extensive testing of both film and HD footage shot with different types of HD cameras led filmmakers to opt for using Sony's HDW-F950 4:4:4 camera for all cave sequences.
"After doing all the tests, basically, logistics of where we were shootingin a cave, where it would be very hard and slow us down greatly to continually be changing film magazinesimage and color fidelity, and the way the HD imagery looked for the underwater sequences were the reasons [the filmmakers] chose to shoot HD for the cave sequences," explains Theodorakis.
But instead of using the cameras as camcorders, Theodorakis engineered an onsite recording approach to HDCAM-SR tape via a Sony SRW-5000 studio deck that was anything but typical, given the inhospitable environment.
At the cave location in the Mexican jungle, Theodorakis says the approach essentially involved using just the camera head of the Sony camera in an Amphibico underwater camera housing, and linking the camera via 2,000ft. of modified Mohawk fiber-optic cable to the SRW-5000 inside a mobile video village located just outside the entrance to the cavean engineering station that featured two portable generators and air conditioning. The same setup was replicated in Romania, several feet to the right of the giant tank.
From the SRW-5000, the signal was downconverted into a Sony DSR-1500A DVCAM deck in order to ingest the imagery directly into an Avid Xpress Pro system for editor Brian Berdan to use for onsite editing. Simultaneously, the DVCAM deck was also used to FireWire the downconverted signal into Theodorakis' Apple Powerbook, where he used Apple Final Cut Pro and DVD Studio Pro to create DVD dailies. From there, those dailies were uploaded to production company Lakeshore Entertainment, Sonythe studio distributing the filmand other collaborators using Apple's iChat system.
But Theodorakis also performed first-pass color correction on the imagery in-camera from his video village station, permitting Hunt to get a good look at what he was shooting inside the cave. According to Theodorakis, that process would also later speed up the digital intermediate process, which was performed at Company 3, Hollywood.
"I had a [Sony BVM-D24 24in.] CRT monitor and a [Sony RMB-750] Paintbox system for setting up the camera for color correction," he explains. "The director watched everything at his own station [in the video village] on an Apple [23in.] LCD Cinema Display monitorviewing the HD signal converted into a DVI signal using [ECinema Systems'] ECinema box. I used a [Sony HDCU-950] camera control unit, which, over the single fiber cable, gave us input and output for com systems to communicate with the operators and actors in the water, return video, and dual link 4:4:4, power, audio, and format-converted 4:2:2 outputs.
"I did the color correct in-camera after doing many chart tests, mostly using the DSC Labs ChromaDuMonde color test chart and various framing and focus charts, as we were framing this for a 2.40:1 aspect ratio," he adds. "Up front, it was a matter of making sure the camera is set to absolute values, which basically means, that I had to ensure the camera had no bias whatsoever. From there, I began the color correction process by painting into the various correction circuits, including the white and black gains, gammas, black gammas, master pedestal [black], user matrixes [16-way color correction circuits], multi-matrix [individual color adjustments], highlight compression, detail adjustment, etc. The key thing to remember is to never go so far with the correction that you cannot take it back later in post. What's more, I monitor the image closely so that I know if I am ever losing any information light and color-wise. Sometimes, I found it necessary to advise the lighting team in the cave to light certain scenes brighter than they normally would."
Theodorakis adds that this kind of "remote engineering" would not have been possible had he not included a Leader 5700 Waveform Monitor/Vectorscope and a Leader 5152DA Waveform Vectorscope in his toolkit. The 5700 scope was used to monitor all signals generally, and the 5152DA scope was used to monitor a 4:2:2 Vectorscope signal coming out of the 5000 deck.
"The Vectorscopes were my most vital tool, to be honest," he says. "They told me everything that was going on in the picture, which was important because the monitors can lie to you depending on your eye's own color bias, conditions, and the inherent color temperature bias of the monitor, among other things."
In addition, he used the scope's still store feature to take still photos from each shot of the waveform readings, the vector readings, and the actual picture itself. That was a crucial step, according to Theodorakis, because imagery captured in Romania had to be blended with imagery captured in the Mexican cave eventually. Light, the color of the water, visibility, and a host of other factors were not identical in those two locations, but having the still photos and waveform data from each location available allowed filmmakers to seamlessly match settings and color information from the different locations.
All of these steps allowed filmmakers to proceed with confidence through the later DI phase, Theodorakis says.
"The whole point was to protect for the DI, and not go so far that we couldn't change it during the DI, but by being so careful, we got the imagery for those scenes about 90 percent of where they needed to go, with the DP present, on set," he says. "Why not get the color on set with the DP's fingerprints all over it? This was a good approach for this project, and later, during the DI, we noticed the color of the water that I dialed in matched across the board [between Romania and Mexico]. We might have timed out lights that got into shots, things like that, and of course, worked to match the film [above-ground scenes were shot 35mm] and HD elements, but overall, we were able to ensure the DI went smoothly and efficiently by taking these remote engineering steps during production."
Theodorakis offers two other hints for folks who might be attempting to drag video equipment into similar environments for feature films. The first suggestion: take durable fiber cable with you, and lots of it.
"We used Mohawk's cable, and Amphibico modified it for underwater use for us, and it was extremely durable, which was obviously important, considering we were dragging it through a cave," he says. "And that one cable was carrying not only the video and audio signals, but actual power, and our com signal, so the director could communicate with the people in the cave, and a lot more."
And finally, Theodorakis insists filmmakers faced "about 100 percent humidity" in both Mexico and on the Romanian stage with the giant water tank. He says they picked up a trick along the way about how to prevent humidity errors in the recording deck.
"One day, when we were trying to record at the tank, we saw the sensor tell us the machine would not record," he says. "We racked our brains and came up with the idea of getting a tank of nitrogenwhich, of course, is a dry, dense gasand every night, when packing up the deck in its case, we blew nitrogen into the case itselfnot the deck, just the case. We never got another humidity error again through the entire production."
Continue the discussion on “Crosstalk” the Millimeter Forum.