The idea behind the little Lytro camera's magic has been around for 20 years or so, but this is the first time we've seen it in a commercial camera. And it could change digital photography dramatically.
Lytro founder Dr. Ren Ng wrote his Stanford PhD thesis on photographing light rays, and it occurred to him that much of this functionality could be shrunk into a small camera. So while the idea of capturing the entire light field, which is what gives the little Lytro camera its magic, has been around for 20 years or so, this is the first time we've seen it in a commercial camera. And it could well change digital photography dramatically.
The result is that photographers can focus after the fact; and those people viewing pictures online can actually interact with the pictures, changing the focus to suit their liking and to create a variety of surprising outcomes.
But really the Lytro promises to change the way people take pictures, making depth of field easier to achieve with a little practice and creativity. At $399 (for the 8 GB version; $499 for the 16 GB one,) this amazing camera could be a huge hit. It won't be available until the the second quarter, but the company is taking pre-orders now.
We got a first hand look at the Lytro camera, trying it out for fun in Las Vegas at CES. And we had a chance to talk to Lytro's director of photography, Eric Cheng, who has created some stunning interaction images on the web site.
The camera was incredibly simple to use (part of the point), but getting it to do the things Eric has created is something that requires practice, an understanding of what the camera is capable of doing, and a few moments of discovery--probably past the time we were given to play with the camera. However Lytro is adding a new feature to the shipping version of Lytro, called "advanced light field mode," which Cheng showed off for the first time, that gave me that sudden "aha" experience. More on that in a moment.
The way the camera works is to capture every ray of light that is traveling in space, at every point in space, and the captured image retains the directional information of the light. As Cheng explained it, as two light rays come into a traditional camera, they don't ever meet, which creates a blur. High-end cameras with heavy-duty sensors do their best to bring those rays together. Because Lytro keeps the directional information, it can detect where the light rays might meet and virtualize the sensor position to make them do so, popping the picture (or whatever part of a picture) into focus. In other words, it is using data to reconstruct a picture after the fact.
Back to the advanced light field mode, then: by selecting this mode, the photographer can choose something in the picture, and make it the center of refocus. In practical terms, if you zoom into an object to make it blurry, or select something in the picture that is blurry already (typically something very close to the lens,) you can tap on that to focus it, and then that becomes the center of everything else you would refocus after the fact.
This lets you create depth of field pictures where one particular object pops out. The idea is to get people to think in terms of 3D, to think about the end experience of someone interacting with the picture. Lytro will have 3D in the not-too-distant future, as well as parallax or perspective shifted views.
The camera can zoom at 8X with a simple swipe of the finger; you can swipe the view finder to move between pictures, and even zoom in and out by simply tapping the image. There isn't really a "pixel" oriented resolution for the camera (like 10 MP); Lytro says it's 11 "mega ray" which is a mixture of pixel information and light rays.
The output is a Lytro image (a .LFP file) and it only works, for now, with Lytro's software. Hopefully that will change quickly. And hopefully we'll see some Lytro accessories, like a tripod or monopods. Cheng said that printouts look good, but only at up to 5x7. See the camera in action in our video, below.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?