With five megapixels and up, today’s digital cameras have more than enough resolution to take photos that can be enlarged to poster size. Yet, to meet demand from customers, manufacturers are racing to cram more and more megapixels onto cameras’ image sensors. But they’re also beginning to consider other ways to distinguish their products.
Now Dalsa, an image sensor maker based in Eindhoven, Netherlands, has developed technology that uses excess pixels to take pictures that look clear and bright – even when there’s just one-quarter as much natural light. Ordinarily, increasing a camera sensor’s light sensitivity introduces more “noise” into an image, in the form of random specks of color. But Dalsa’s technology, described at this week’s meeting of the Institute of Electrical and Electronics Engineers in Washington, DC, increases sensitivity without adding noise.
Dalsa’s sensor, which is designed for professional photographers, has a capacity of 28 megapixels – or more than five times that of a typical consumer digital camera. But this impressive figure isn’t noteworthy in itself. This fall, Kodak announced 31- and 39-megapixel sensors, for example, and Dalsa has made its own 66-megapixel sensor, although not for mass production.
“At this point everybody knows how to put many pixels on a piece of silicon,” says Albert Theuwissen, Dalsa’s chief technology officer. “Differentiation comes through features.”
What’s new is what photographers can do with Dalsa’s technology. For one thing, they can record from each pixel separately, creating a gargantuan image with enough information to produce sharp, wall-sized ads. Or they can combine signals from clusters of neighboring pixels, in effect amplifying the signal while averaging out the random noise.
This latter technique, called “charge binning,” because the electronic charge from multiple pixels accumulated as if in bins, has been used in the past for black and white photography. But adjacent cells in a color sensor alternate between green, red, and blue, in a checkerboard pattern, so that simply adding up the light readings from these neighboring cells would blur the colors together. Dalsa’s innovation is a system that combines signals only from the same-colored pixels, keeping the colors distinct. Theuwissen claims that Dalsa is the first to have accomplished this.
In addition to increasing light sensitivity, combining the pixels has the benefit of improving what’s called “dynamic range,” a camera’s ability to capture images that have both bright and dark areas, without losing detail. As sensor makers have stuffed more and more pixels into small point-and-shoot cameras, they’ve also made each individual pixel smaller, actually limiting the range of light levels they can distinguish. In effect, charge binning makes the pixels larger, creating “virtual pixels” that are able to absorb more photons and distinguish a greater range of light levels.
Achieving this greater dynamic range and sensitivity does come at the cost of some resolution. Since the Dalsa sensor adds up information from groups of four like-colored pixels, its output resolution is only seven megapixels, instead of 28. But seven megapixels is still enough for pictures that run in print magazines, according to Christopher Harting, a long-time professional photographer (and an occasional contributor to Technology Review).
Theuwissen says the extra light sensitivity provided by charge binning would be useful in situations such as sports photography, where high shutter speeds are needed to freeze the action. (The faster the shutter speed, the less light that reaches the sensor.) A photographer using the sensor would have the option of switching between modes in different situations.
But Harting points out that medium-format cameras, like those the Dalsa sensor is designed for, are typically used in situations where photographers have control of the lighting anyway, such as in studios, and where there isn’t a need for high light sensitivity. In contrast, sports photographers tend to use more smaller-format, portable cameras, he says. Dalsa’s success with its new sensor, therefore, may depend on whether medium-format camera users find applications for the switchable sensor.
Other companies are also looking at ways to distinguish their cameras by improving low-light sensitivity. Fujifilm, for instance, has developed computational methods for cleaning up the noise in an image.
Theuwissen says Dalsa plans to keep working on highly-sensitive, low-noise sensors, while looking for other tasks to accomplish with extra pixels. Such technology could eventually be used in consumer cameras – giving them a better reason to be so musclebound in megapixels.
The big new idea for making self-driving cars that can go anywhere
The mainstream approach to driverless cars is slow and difficult. These startups think going all-in on AI will get there faster.
Inside Charm Industrial’s big bet on corn stalks for carbon removal
The startup used plant matter and bio-oil to sequester thousands of tons of carbon. The question now is how reliable, scalable, and economical this approach will prove.
The dark secret behind those cute AI-generated animal images
Google Brain has revealed its own image-making AI, called Imagen. But don't expect to see anything that isn't wholesome.
The hype around DeepMind’s new AI model misses what’s actually cool about it
Some worry that the chatter about these tools is doing the whole field a disservice.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.