Lytro, a new kind of photography?

Photography has certainly come a long way from the days of glass plates and film. Now a new company, Lytro, Inc. thinks that they're on the verge of creating "Camera 3.0" with their soon to be released "light field" camera later this year. From reading their press release and checking out their sample gallery the idea is very intriguing.

The concept is that you really don't have to decide where to focus the camera image until after you've made it. The idea is that you shoot your shot and then you realize that you "focused" on the wrong subject afterwards and you then go to your computer and "refocus" on something else in the image. No more shots of walls in focus and your son out of focus, you just go back in and change the point of focus and you're good to go. Again, a very interesting concept.

Other features according to the release are "unparalleled speed", because you don't have to worry about waiting for the auto-focus to find your subject and do its job.

Then there is the social media aspect. It seems all good new concepts must have these. There's is tagged as "Living Pictures". It allows viewers to play with your images without the need to download software. I did note that to see the sample images Adobe's Flash was required so maybe you won't be able to view these images on Apple's iPad.

Another aspect is that it uses all the available light in a scene giving it "remarkably low light" sensitivity without a flash. Also, it features "Immersive 3D" which is said to go "beyond conventional stereo 3D by, for example, controlling the perspective view of a scene".

Okay, it all sounds great. And I can't say that I'm surprised at all. For the moment I haven't had a chance to read Dr. Ng's dissertation on how this all works, but I have seen the application of software and fast processors in the latest cameras that give me some hints as to how this all might work.

Take the line of Sony cameras that were introduced with the help of Taylor Swift two years ago. One of their features was the iSweep Panorama. It allowed the user to start the shot and turn the camera capturing almost 180 degrees as it took about 100 individual images and then spliced them together into a single shot that was smart enough to recognize faces and not chop them up. Or their Motion Blur reduction mode that takes multiple shots at very fast shutter speeds and then combines them to build up an image that couldn't be captured with just a single shot in marginable light.

Here's my confablulation on how I'd do this if someone gave me enough money to find smart folks to do the actual work. I'd design a camera that would act like a pin hole camera that would have a very small aperture that would give me a lot of depth of field making everything from right infront of the camera to infinity look sharp. Then using the Sony model shoot a lot of images as quickly as possible and combine them to make a sharp image...then figure out what to selectively defocus parts of the image and apply that. With even more programming figure a way to tell the image where to shift what's in focus.

Another thought would be to again shoot a lot of images while changing the focus from near to far and store them together until the user decides where they want to focus.

Obviously, the task isn't easy, but it seems these folks have it figured out. Of course for the purists in the crowd, the solution is to think about what you're doing in advance and compose and focus your image accordingly, but heck that's Camera 1.0.

Dave Dornlas