Dylan Tweney

New Chips Poised to Revolutionize Photography, Film

For the first time, professional-grade single-lens reflex cameras are gaining the ability to record high-definition video. That capability, photographers say, has the potential to transform both still photography and moviemaking — and it’s largely thanks to advances in the semiconductor technology u
Dylan Tweney 6 min read

For the first time, professional-grade single-lens reflex cameras are gaining the ability to record high-definition video. That capability, photographers say, has the potential to transform both still photography and moviemaking — and it’s largely thanks to advances in the semiconductor technology used to make the image sensors inside these cameras.

"I think this is the holy grail for news photography," says Randall Greenwell, the director of photography for the Virginian-Pilot, a newspaper in Virginia.

Greenwell says photojournalists are already shooting both stills and video, but using separate equipment for each medium, which is awkward, cumbersome and requires additional training. With a single camera that can do both stills and video, he says, the job of the new-media journalist will be greatly simplified.

"With that kind of flexibility, it’s going to be a real game changer," Greenwell says.

While compact digital cameras have had video-recording capabilities for years, the image quality provided by these cameras has been disappointing because of their small image sensors and comparatively poor, miniaturized optics. High-end video and movie cameras produce top-notch HD video and their interchangeable lenses give filmmakers the creative control they crave, but the cameras are big and expensive. Even the RED ONE, a super-high-definition movie camera that records digital video that’s comparable in quality to that of film stock, rings up at about $17,000. That’s a bargain compared to movie cameras, but it’s still a lot of dough for most people.

By contrast, the 21-megapixel Canon 5D Mark II, which shoots 1080p HD video, will cost $2,700 (plus the cost of lenses) when it becomes available later this year. The 12-megapixel, highly rated Nikon D90, which records 720p HD video and is available now, costs even less: a mere $1,300 gets you the body plus a basic zoom lens.

Both cameras deliver extremely high visual quality for both still and moving images — and just as important, they allow photographers to use a wide complement of interchangeable lenses, from macro lenses for extreme closeups of insects to long telephoto lenses for shots of offensive plays on the other end of the football field. That’s important to pro photographers, for whom lens choice is a critical component of the creative process.

"The single biggest difference between still photography and a movie, aside from motion, is lens choice and depth of field," says Vincent Laforet, a Pulitzer Prize-winning photographer who is part of a Canon marketing program, "Explorers of Light."


Laforet also touts the Canon’s ability to capture images when there’s not much light, an impression confirmed by other photographers. "That you can actually capture in available light is going to be a big difference," says Greenwell.

Laforet predicts that this low-light sensitivity will lead moviemakers to dispense with expensive, bulky, and obtrusive lighting equipment, shooting their movies entirely with available light.

In addition, the new cameras are small compared to professional video cameras, enabling photographers to shoot in a variety of situations with relative ease. Laforet, for instance, shot a demonstration video using the Canon camera over the course of a weekend, incorporating shots that required him to lean out of the open door of a helicopter.

The key to the Nikon’s and Canon’s incredible image quality lies with the large image sensors they contain. Whereas a typical compact camera might have an image sensor measuring about 5mm by 7mm, the sensor on a "full frame" SLR like the Canon 5D Mark II is the same size as a frame of standard camera film: 24mm by 36mm. That’s a more than 24-fold increase in image area. (The Nikon D90 uses a smaller 16mm by 24mm sensor, but even that is 11 times the area of a compact camera’s imaging chip.)


The increased size of the SLR’s sensor allows each individual pixel to be larger, reducing the amount of "noise" in the image and increasing the amount of light each pixel is able to capture. The result: Dramatically better images, even at the same or lower number of megapixels, especially in low light.

A larger sensor also means it’s easier for photographers to control the depth of field. Compact cameras have short focal-length lenses to match their small sensors. The laws of optics dictate that these lenses have a large depth of field.

"As image sensor size decreases, effectively you are getting more and more depth of field," says Chuck Westfall, a technical advisor at Canon. For point-and-shoot cameras, that’s convenient, because it’s harder to get an accidentally out-of-focus snapshot. But for creative photography, being able to control the depth of field is essential. That’s how you get those portraits where a person’s face is sharply in focus, while the background is pleasantly blurred.

So why has it taken so long for digital SLRs to add video-recording capabilities? The answer has to do partly with the physical design of SLRs, and partly with the type of imaging chips used.

Inside every SLR is a flip-up mirror that directs light either to the viewfinder or to the image sensor, but not both at the same time. In order to record video (or provide a live image on the LCD), the camera has to "lock up" the mirror, blocking the viewfinder. The pros who until recently defined the digital SLR market were initially loathe to do that because of the better optical quality afforded by the viewfinder.

"The viewfinder is arguably the best way to see your picture as you compose it, and it also offers the best, most stable platform for shooting SLR pictures," says Steve Heiner, a senior technical manager at Nikon.

But perhaps the most critical component of the new generation of cameras is the imaging chips inside.

For most of the past decade, consumer cameras have used a kind of imaging chip technology known as charge-coupled device (CCD). Recently, a competing imaging technology known as complementary metal-oxide semiconductor (CMOS) has come to the fore, largely because of its lower power requirements. CMOS chips appeared first in SLR cameras aimed at the high end of the market and have only recently started appearing in point-and-shoot cameras, which are still dominated by CCD technology. What drove the transition to CMOS was the large sensor size of SLRs.

"The power consumption of a CMOS is so much lower [than CCD] at the full frame size that this is the only way you could come up with a reasonable battery life," says Westfall.

But CMOS chips initially had trouble delivering live video images due to overheating, the need to come up with a way of resampling images on the fly (converting them from the sensor’s maximum capacity to the smaller resolution of HD video) and other problems.

It wasn’t until 2006 that Olympus first offered a digital SLR with a "live view" option, which kept the imaging chip in constant use while delivering a live image to the LCD. The feature proved popular, and other manufacturers soon followed suit.

Once they’d added live view, it was a small step for manufacturers to add the ability to record the video coming off the sensor instead of merely directing it to the screen on the back of the camera.

Now, experts say, CMOS imaging technology is developing much faster than CCD, partly because CMOS imaging chips are built with the same basic processes used in producing other kinds of semiconductors, like memory chips and processors. CCDs, by contrast, are less familiar to the majority of semiconductor engineers.

And thanks to Moore’s Law, the power and speed of semiconductor technology keeps increasing exponentially. That means CMOS image sensors are getting better and better, incorporating more sophisticated noise compensation, shrinking the size of the gaps between each light-gathering pixel that are devoted to wiring and other electronics, and adding image and video processing features to the chips themselves.

"I’m amazed myself at how quickly the tech developed a life of its own and how fast it’s evolving," says Eric Fossum, an entrepreneur and engineer who developed the type of CMOS imaging technology used in most modern cameras while he was a researcher at NASA’s Jet Propulsion Laboratory in the early 1990s. "It’s kind of mind blowing to me."

Link: New Chips Poised to Revolutionize Photography, Film

Link broken? Try the Wayback Machine.

More from Dylan Tweney

How Microsoft Researchers Might Invent a Holodeck

My final story for Wired.com appeared August 31. It’s a look at some of the work that the scientists and engineers in Microsoft’s research division are doing to create the computer interface — and communications screens — of the future. REDMOND, Washington — Deep inside Microsoft is the brain of a m
Dylan Tweney 2 min read

Infiniti Hybrid Is a Green Sedan for Silver Foxes

If you’re old enough to remember the energy policies of the Carter administration, green enough to have donated to the Nature Conservancy and young enough to get a rush of testosterone from dusting that polo-shirt-wearing jerk in his BMW, Nissan has the car for you. And though its styling walks a fi
Dylan Tweney 1 min read

How to Make a Clock Run for 10,000 Years

High on a rocky ridge in the desert, nestled among the brush, is the topmost part of a clock that has been ticking for thousands of years. It looks out over the ruins of a spaceport, built by a rich man whose name was forgotten long ago.
Dylan Tweney 28 min read


Subscribe to my newsletter on writing & storytelling

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Dylan Tweney.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.