Behind the Scenes at National Geographic

It's like watching an underwater chase scene in a James Bond film. The camera dives, somersaults, whirls, barrel rolls, plunges to the bottom, then shows a glimpse of the silvery underside of the water surface before knocking opponents out of the way.

But what you are looking at is not a Hollywood movie. Instead, it's a documentary showing a seal's eye view of the world, made by a waterproof camera strapped to a wild seal's back as it chased after its prey. Welcome to Crittercam - possibly the smallest, wholly integrated autonomous underwater vehicle system.

The basic Crittercam is a 2-pound, 3.5-inch-wide, 10-inch-long torpedo-shaped aluminum cylinder with tiny, streamlined lights and microphones built into the outside, and a hemispheric glass optic on the front. Inside is everything a movie-maker or a scientist needs: batteries, audio equipment, CCD chips and image intensifiers. Additional devices simultaneously record depth, time, duration of dive, temperature, and velocity. (Some versions are coming out that also measure acceleration, light level and compass orientation.) A beacon gives satellite telemetry. The cylinder can withstand depths of 7000 feet, and it is large enough to hold six hours of Hi-8 or digital videotape. An on-board micro-processor is programmed to switch the camera on and off at the researcher's discretion, shooting the whole tape in one session or in brief segments spread out over five days.

And it does all this without harming the creature or interfering with its behavior, said Crittercam's inventor, Greg Marshall of National Geographic Television in Washington. The system is designed so it does not increase the animal's hydrodynamic drag, and it is just buoyant enough to bob along without affecting its host. "The camera is just a hitchhiker along for the ride. Animals get used to it remarkably quickly and swim away after a few minutes," said Marshall. Crittercam comes close to approaching the ultimate goal of documentary photography - a camera floating unnoticed in space, taking pictures without influencing its subject.

It's a completely different approach from the bigger, heavier remotely operated vehicles, whose power cords and fiber optic cables tether them to a mother ship like a dog on a leash. In contrast, no wires or cables of any kind connect battery-operated, autonomous underwater vehicles like Crittercam to a mother ship. In fact, there is no mother ship. These devices are completely free and independent. Researchers have no control over where they go or what they snap pictures of.
The conditions sound impossible. And the systems still have to contend with many of the same problems as their tethered counterparts, including high pressure, little or no light, temperature extremes and leaks.

But in the course of 13 years of experimentation, Marshall and his team have successfully made more than 300 dives with different versions of Crittercam, taping the daily lives of over 28 species from walruses to whales. There are currently 20 Crittercam systems in existence, available to scientists for pure research. While a still frame grabbed from a Crittercam is not quite as sharp as one from a remotely operated vehicle equipped with high-definition television, Crittercam offers a unique porthole into a previously inaccessible world. Crittercam is changing our understanding of the undersea world in much the same way as the first scuba gear.

Marshall cited several behaviors first identified by Crittercam. In an article published in the April 2000 issue of Marine Mammal Science, researchers described how they used it to find the likely cause of the decline in the Hawaiian monk seal population: The animals may be starving to death. The camera showed that these seals do not feed in the shallow, legally protected reef zones as previously thought, but instead hunt in deeper, unprotected waters where commercial fishing boats can out-compete them.

Crittercam also showed how seals and whales submerge - by gliding instead of swimming. They stroke vigorously for the first 30 meters as they head down, then cruise the rest of the way, as their bodies compress to become negatively buoyant and sink. The huge energy savings helps explain how air-breathing creatures can stay under water for lengthy periods.

And researchers have tried putting the camera on one species - sperm whales - to spy on another species - the seldom-seen giant squid. Clyde Roper, a zoologist at the National Museum of Natural History in Washington, said that because sperm whales are known to hunt squid, they can be used like hunting dogs to locate and film the creatures. The concept is an unusual twist, he said in the Geographic's subsequent film Sea Monsters: Search for the Giant Squid. "We really don't know much about what happens to (sperm) whales once they leave the surface," Roper said. "So we're working with a mystery that is hunting a mystery."

The technology itself may seem mysterious at first to outsiders, but Marshall said that basically he and his team start with existing technology - such as the streamlined aluminum tube and the hemispheric optic - and modify it. "We essentially strip a video camera down and make it mimic a remora," he said.

One of their biggest challenges was acquiring enough light for the camera to capture an image in the dark depths. Because powerful underwater lights would disrupt the behavior of the very creatures they were trying to video, they chose to install image intensifiers. The technology, made by ITT Industries Night Vision of Roanoke, Va., was originally designed for military use to improve devices such as sniper scopes and binoculars. Marshall found that these devices could amplify the existing underwater light by a factor of 50,000, allowing cameras to be used as deep as 600 meters without artificial lights.

Beyond those depths, Crittercam uses near-IR LEDs that have been specially tweaked by the Geographic's technicians to emit light at a wavelength that can be picked up by the video camera while remaining invisible to marine creatures. (Like many researchers, Marshall did not wish to go into details about the wavelengths used or how the equipment was modified, except to say that they start with standard off-the-shelf components from Hewlett-Packard.) These LEDs project light 2 to 3 meters out in front of the camera.

He said that knowing when to use image intensifiers and when to use LEDs is a matter of experience with the habits of the creature they are trying to film and the local environment. They discovered that when penguins equipped with Crittercam go just 2 meters under Antarctic ice, the light disappears quickly; "Next time we'll use LEDs," Marshall said.

Their two other big problems were relatively low-tech, but hard to solve: How to attach the camera system to the animal? And how to retrieve the video afterward? Seals were relatively easy, because they periodically come out of the water, returning to the same place ashore each time. Scientists could re-capture them, remove the Crittercam and extract the used videotape. But the situation is different with tiger sharks, which are hard enough to catch the first time. And the researchers must collect the shot video; "Unless we retrieve it, it's a bust," said Marshall.

After much trial and error, they solved the attachment problem with customized mounts that physically bind the camera system to a specific maritime critter. They use suction-cups for whales, harnesses for dolphins, fin clamps for sharks, soluble epoxy patches for seals and hose clamps for walruses.

They solved the problem of retrieving the videotape by incorporating a specially designed pin into the mount; The pin dissolves after a predetermined time in the water. Once it disintegrates, the mount falls apart and sinks to the bottom, allowing the buoyant camera system to float to the surface. A radio beacon inside enables the researchers to home in on the drifting Crittercam and scoop it up.

Marshall said that he would like to improve the retrieval of the videotaped images. Rather than plucking a used Crittercam from the open sea, he'd like to have some way of electronically transmitting the video directly from the camera to a shipboard viewer in real time, without any wires or cables dangling in-between. "That's something I was hoping your readers could answer," he commented.

His team is also looking for brighter LEDs, more efficient image intensifiers and more sensitive CCD chips.
Whatever they use has to be all solid-state and very rugged, Marshall stated. "You're on the back of a wild animal," he commented. "A great white shark can be pretty tough on this equipment."

Λ back to top