When a set of online teasers for a new camera called the Lytro appeared earlier this year, you could have been forgiven for seeing the invention as just another gimmick. The camera’s attention-grabbing feature is a kind of after-the-fact autofocus: with a click, any blurry portion in a picture can be snapped into sharpness—another step in the march of idiot-proof photography. [...]
The underlying technique is called “light-field photography.” A traditional camera, of course, captures light reflected off its subject through a lens and onto a flat surface. Proper focus is important to ensure that the image you get is the precise slice of visual reality you want. But “computational photography,” pioneered by Marc Levoy of Stanford University and others, takes a different approach, essentially using hundreds of cameras to capture all the visual information in a scene and processing the results into a many-layered digital object. One of Levoy’s former students, Ren Ng, added the twist that resulted in the Lytro: instead of using multiple cameras, he integrated hundreds of micro lenses into a single device.
The upshot is a photograph that’s less a slice of visual information than a cube, from which you can choose whichever layer would make the most pleasing two-dimensional image for printing and framing. But you can also leave the picture as it is—a three-dimensional capture suitable for digital display or distribution—and let others do the fiddling. Rather than a definitive, static image, a light-field visual object is intrinsically interactive.
Not mentioned, but apparent, by combining with Photoshop, you could select parts from different layers and merge into one image, no doubt achieving some startling effects. Yes, it's more idiot proofing, but so what? You still need an eye for light and composition and a good photographer will still do better than a rank amateur like me, even with all that technology.