The Sharpest Image

The man behind the world's most powerful camera confronts killer viruses, nude sunbathers and the San Diego Padres

Share

Graham Flint is the sort of man who uses the structural bracing of a nuclear reactor’s safety door as a camera stand. The bracing secures his camera casing to the inside of his minivan and is indicative of the precision and focus with which he approaches all aspects of his life, none more so than his current and most ambitious project: a 1,000-shot survey of America at the dawn of the 21st century, his Portrait of America, taken with the camera he designed and built, the highest-resolution landscape camera ever created.

It’s surprising then, when he announces that he’s forgotten his calculator. We’re standing in front of Albuquerque’s San Felipe de Neri Church, the oldest Catholic parish in America and the last stop on the last leg of his journey. I offer the one on my cellphone, but it won’t do. “I need a scientific calculator, with trig functions,” he says. Undaunted, Flint whips out a small notebook and begins jotting down a long series of calculations.

Flint is crunching numbers to help him focus his shot. His camera, for all its custom-built sophistication, has no viewfinder. The reason it has no viewfinder is that its magazine-which holds huge 9-by-18-inch frames of Kodak film, at $1,200 a roll-comes from a Fairchild Aviation Corporation K38 aerial camera, used in high-altitude reconnaissance flights at the height of the Cold War, in the mid-1950s and -’60s. It was never meant to take landscape photographs-until, that is, Flint decided it could.

Dressed in a khaki vest and a floppy safari hat, with owlish glasses and a posh English accent, Flint fairly invites gawkers. Several ask if he is filming a movie. What he is actually doing is preparing to take a shot that will put the tourists’ high-tech, autofocusing digital cameras to shame: a photograph that will approach four gigapixels in resolution, or 1,000 times the capability of the average Elph or Coolpix. Flint’s images offer not only a seductive glimpse into the future of photography, the promise of incredibly large and detailed images; they are also providing an invaluable record, as never seen before, of the way America looks now, its landscapes and monuments preserved for posterity with breathtaking clarity and scope.

How did this semi-retired physicist and pioneering laser-weapons researcher, working with an analog camera whose parts date to the Eisenhower administration, come to reinvent photography in the early 21st century, making images that are at least a decade ahead, in terms of resolution, of the most advanced digital cameras on the market today? It all began, rather improbably, with mice and the Milky Way.

The magazine [1], which holds 9-by-18-inch film, was formerly used in U2 spy planes. Graham Flint converted it to landscape use by grinding a custom lens [2] out of six types of glass. He focuses his shots manually, with knobs [3] that can align the lens along three axes.

The Road to 1,000 Megapixels

To create analog gigapixel photography, while digital photography was still in its megapixel infancy (“mega” means million; “giga,” billion), virtually required someone like Graham Flint: a hobbyist photographer with an extensive knowledge of optics, physics, astronomy, and military aerial-reconnaissance cameras. After graduating in physics from the University of Birmingham in England, Flint emigrated to Seattle in 1958 to work on high-altitude remote sensing for Boeing, setting into motion a long career studying how light moves through air. After his five-year stint at Boeing, he left to head the Laser Devices Laboratory at Martin Marietta, then set off to co-found his own company, International Laser Systems, which he sold in 1983. Since then he has served as director of the U.S. Air Force’s Developmental Optics Facility and chief technical officer of the Laser Power Corporation, and has designed and built a prototype digital camera for the Hubble Space Telescope. These days, he still acts as a consultant on various projects, commuting regularly to California from New Mexico.

But that picture, however impressive, is incomplete. Flint is a man of dizzyingly diverse talents and pursuits, what he calls his “avocational” side. An amateur architect, he designed an Italianate villa for himself-“Hadrian West”-atop Sandia Mountain in New Mexico; he also designed the classically inspired room in his current house in which he and his wife, Catherine Aves, were married in 1990. He has dabbled in antiques, historic restoration, philately (with a particular interest in British Empire stamps under King George V) and rare-book collecting, and he once restored to racing condition a Porsche RS60 he found in the service area of a Florida Porsche dealership. Only 14 of the cars were ever made.

In addition to his career and multitudinous avocations, Flint found the time to develop a telescope that searches for planets in other solar systems, which he placed in the observatory of Hadrian West. And in the late 1980s he began work on the “Vista Galactica” project for the Joint Observatory for Cometary Research in Socorro, New Mexico (now the New Mexico Tech Remote Observatory), a complete, high-resolution (in the one-gigapixel range) photographic survey of the Milky Way. Flint built a 6,000-pound wide-angle camera in his garage. It was so large that to move it out of the garage to its eventual home on New Mexico’s South Baldy Mountain, he had to have one side of his driveway lowered. He faced a more serious hurdle at the Observatory, however: the discovery of hantavirus, a lethal disease spread by rodent droppings. The facility was ordered closed, and Vista Galactica was abandoned.

By the late 1990s, Flint had officially retired, but, he says, he “got bored real quick.” Soon came what he describes as a recurring itch. “Most of my career I’ve spent applying physics to various problems. Once every 10 years, I suppose, I ask myself: Within the realm of the technology I’m working in, what remains that hasn’t ever been done?” Flint was also looking for a project that he and his wife could tackle by themselves. “We said, ‘Let’s do something where everything is within our control, where we can put our instruments in a van.’ ” That’s when Flint, long oriented to looking at the heavens, decided instead to focus on the world around him.

The Digital Divide

To look at a photograph in the multiple-gigapixel range, I learned in an afternoon with Flint at his dining-room table, is to see the world in a new way. From a few feet away, a panoramic image he’s taken of the Grand Canyon is startlingly vivid, almost hyperreal. As I move in closer, I am amazed that it retains its clarity. It does not become pixelated. Where I saw only a distant group of visitors standing across the canyon, I now see the fine chain-link fencing behind which they stand. Flint’s photographs are like the hard-to-believe technology in spy thrillers-it turns out that you can, beginning from a panoramic shot of San Diego’s skyline at night taken from a quarter of a mile away, keep zooming in until you can make out the poster on an apartment wall.

Flint turns to one photograph that shows the staggering potential of gigapixel photography. The shot is of a recent baseball game, White Sox versus Padres, taken at San Diego’s Petco Park. Flint cajoled the stadium officials into letting him set up his camera in center field, in the section kept empty so that
batters can see the ball. His photo looks straight out of Sports Illustrated-at first. Then you realize that what might
normally be an anonymous crowd of tens of thousands blurred in the background is not so anonymous. “I could get a passport-quality headshot of everyone in the picture,” Flint says. The photograph also captures the ball, hovering in the batter’s box, about to be hit for a three-run home run that would give the Padres the lead. When Flint makes a large-scale print of the image, he’s hoping to be able to make out the stitches on the ball. “It depends if it was spinning,” he says, smiling.

The creation of what are essentially high-altitude surveillance photographs brought down to Earth required intensive engineering. The trick was to make something meant to take photographs at 70,000 feet perform at 200 feet. “Commercial lenses are pretty good, but they’re designed to work with 8-by-10 negatives,” Flint says. “We’re twice that size. You could get a lens that gave good performance at the middle of the image but was fuzzy and soft out at the corners.” Using a set of surplus Fairchild A7 spy-plane magazines he had procured at an auction, he and a friend, Paul Weissman, machined an ultra-wide-angle, extra-sharp lens, dubbed the Asymmagon, that could produce adequate clarity across all of the camera’s 9-by-18-inch negatives.

Although the name for his mission, the Gigapxl Project (“gigapixel” won’t fit on a license plate), implies digital photography, Flint is careful to stress that his camera is entirely analog. Once his film is processed by a lab in Ohio, Flint digitally scans the images using a high-end Heidelberg drum scanner he recently purchased, makes color corrections and other slight modifications using Photoshop, and prints the final product in sheets that are patched together into a single image of up to 8 by 16 feet. “It’s a balance,” Flint says. “There’s no CCD [charge-coupled device, the sensor that acts as the “film” of a digital camera] that can catch this much information, but there’s no film processing that can use this size negative and can go as big as we’re printing. We’ve got to combine the two technologies.” Flint believes that CCD technology will improve but that it has fundamental limits. “When you actually look at the details of a CCD device, each little photosensor has to have some circuitry with it,” he explains. “If we want to have pixels closer together, what happens is that the processing circuitry starts to occupy a larger fraction of the area, and so now the area that’s the sensor gets smaller and becomes sensitive. You squeeze one place, and you lose somewhere else.”

Despite his camera’s meticulous engineering and its ingenious marriage of technologies, Flint demurs that it isn’t particularly cutting-edge. “This whole project is nothing new,” he says. “It’s just having a familiarity with all the different aspects and seeing where we can push them and how we can combine them to get the best end result.” Flint’s self-assessment is disingenuous, of course-he undertook the project precisely because landscape photography of this scale, with this detail, had never been attempted before. But it illustrates something about his outlook: that the hardest part of the project was, in a way, the realization that it could be done. Once that was established, success was a simple matter of insight, hard work and an unyielding application of logical problem solving.

His is a scientist’s view of photography. For Flint, a photograph represents a whole process by which information is lost. Each step along the way, from the air to the lens to the digitization “noise” of the scanner, slightly corrupts the original signal. “What [my wife and I] have done is mathematically analyzed that entire chain,” he says, “so we can predict from the scene to the final output what are all the contributors that convolve together to degrade your original information content. Then we’ve tackled them to make sure there is no weak link in the chain.” Take, for instance, the weather. Flint tends to shoot just after a rainstorm. He explains that “immediately after a storm, the surface of the ground has been cooled. There’s less turbulence from the ground from rising air. You also have fewer aerosols in the atmosphere, so you get less small-angle scatter and you get crisper results.”

Flint is an inveterate tinkerer, a boffin par exemplar. This is evidenced not only in the camera but in the system he created to help shoot the Portrait of America, his sprawling photographic survey of the country. Standing in his driveway, he opens the back of his minivan and pulls out the large housing that holds the camera. It slides out on the aforementioned nuclear-reactor frame. “A pair of those will hold a 2,000-pound door,” remarks Flint, knowing that they will never be pushed anywhere close to such engineering limits. The tripod, also built by Flint-commercial models couldn’t support with the necessary stability the camera’s 80 pounds-does double duty, attaching to the wheeled camera like a wheelbarrow handle for easy transport. It’s all pure Flint: a machine perfectly designed and crafted, from intricate optics down to custom carrying mechanism, his ambitious vision manifest in every thousandth of an inch.

The Road Trip Goes Google
When I met up with Flint, he had just finished the 12,000-mile, 250-shot final leg of the Portrait of America. Flint and Aves’s main task now is to finish the Portrait, fine-tuning the digitized images of purple mountains and fruited plains (in which the fruit can be seen), matching the colors of photographs of the Rocky Mountains by referencing souvenir samples of the actual rock. “We’re very fussy,” Flint says. “We consider that what we’re doing is ‘This is America at the turn of the century.’ And we would like it to stand for all time as a good and accurate documentary.”

Carol McCusker, curator of San Diego’s Museum of Photographic Arts, which recently hosted a show of Flint’s photographs (visitors were given magnifying glasses to delve into the detail), sees in Flint an echo of William Henry Fox Talbot, the 19th-century English scholar and consummate dabbler who is the acknowledged inventor of modern photography. If some question the presence of Flint-who, by his own admission, is a scientist, “not an artist”-in an art museum, McCusker disagrees. “Technology has always driven photography,” she says, expanding the tools the artists have to work with.

As word of the Gigapxl project has begun to spread, Flint has been fielding requests to adapt the technology for one purpose or another. Someone wants to make the world’s largest jigsaw puzzle. An engineer from New York wants to take pictures of buildings so that the state of their brickwork can be studied for insurance purposes.

But one of the most potentially productive calls came from Michael T. Jones, chief technology officer of the aerial-
imaging company Keyhole, which was recently acquired by Google as part of its Google Earth initiative, a searchable image database of the Earth from above. Jones was amazed that someone had taken an aerial camera “in the same broad field that we’re using [satellite imagery]” and fashioned a terrestrial camera. “As soon as I saw Graham’s camera, I was entranced,” he says. “I became Graham’s acolyte.”

Jones and Flint share an aspiration for the next phase of Gigapxl: a high-resolution survey of the world’s endangered monuments. As Jones sees it, the Gigapxl images could serve as a permanent record of sorts. “So if the next Taliban blows up the next Buddhas,” Jones says, “we’ll have recorded it so accurately that people can reconstruct it perfectly.” Jones envisions Google Earth, with its technology and storage capacity, as the eventual repository of the images: “That’s our business, zooming in and out of giant pictures.”

But Flint knows that he may already have devised something more profound than a crisper image. He tells the story of how the 19th-century photographer Eadweard Muybridge-who famously showed that all four of a galloping horse’s hooves are off the ground at the same time-used his camera as a scientist would. “He’s somebody who used photography to create an image that was both startling and different and gave people an understanding that they didn’t have before,” Flint says. “We’re trying to take images that people have photographed thousands of times and give people a whole new perspective of those scenes.” Despite Flint’s own protestations to the contrary, perhaps the 21st century has found Muybridge’s corollary: the
scientist who uses his camera as an artist.

Tom Vanderbilt is an amateur photographer and a contributing editor at ID magazine. He lives in Brooklyn, New York.