Focusing after the shot, the plenoptic camera: Digital Photography Review — Astonishing technology — if it works.
Ren Ng, graduate student at Stanford University has developed a hand-held plenoptic camera which takes a shot first and allows you to make the decision about focus point in software after the event. The prototype camera is actually a Contax 645 with a modified Megavision FB4040 back (sixteen megapixel). The back has had an array of 90,000 microlenses mounted in front of the sensor (with a gap between the array and the sensor). These microlenses create a unique image on the sensor surface which includes not only the amount of light deposited at that location, but how much light arrives along each ray. The image is then reconstructed in software and a focus point can be chosen. Note that the final resolution is the same as the number of microlenses.
I’ll do the math:
300 X 300 = 90,000
“Note that the final resolution is the same as the number of microlenses.”
My conclusion is that it’s not going to find a big market.
Pretty funny. I hope they don’t start adding these to camera phones, or you’ll have a 1000 pixel image.
Interesting to see, given all the complaints about current autofocus systems (needing assist-lights in low light, “hunting” at the telephoto end of a zoom, etc).
If they can scale the micro-lenses to greater pixel-counts, they should have a nice system for computer-based photographers. [It won’t be of any use for people who don’t normally process their pictures on a computer before viewing/printing.]
Interesting idea. I don’t see much use at the consumer level, yet. High end users will find many uses and probably afford the high cost. I picture it being used for 3-D purposes.
American ingenuity at work.
My conclusion is that it’s not going to find a big market.
As if the Contax 645 with a 16-Mp digital back has a large market. 🙂
He found a big-enough existing camera body to fit a prototype micro-lens array – vice trying to build a camera from scratch.
Its only been the last couple of years that consumer digital cameras have gone into the multi-megapixel sizes. If this lense-array can be scaled to even 4-Mp it would then handle most 4×6 snapshot needs – and be that much more “goof proof” than any existing camera.
Exposure and color accuracy are fairly fool-proof now, but auto-focus still needs work, especially for non-SLR designs which can’t place AF sensors in the light-path [ie. Sony’s new R1]
This has a huge potential if you understand the photographic implications. One of the major challenges in ‘macro’ and ‘micro’ photography is depth of field, which is basically how much of your photo is considered in-focus infront and behind the main plane of focus. Having big depth-of-field typically means needing huge amounts of light, sice depth of filed is controlled principally by the F-Sttop of the lens, and a smaller F-Stop means greater depth of field.
Suppose that you can only take one photo of an item, and that you have to compare various depths within the same item. For example, something that is moving, such as an airplane turbine being struck by an object. You don’t need huge resolution, but you do need the ability to see what is happening throught the depth of the turbine. Being able to shift the focus back-and-forth within the same photo is a HUGE development. Medical applications are numerous, as are uses for micro-biology, metallurgy, and almost every other field of science.
Pixel counters just don’t get the importance of being able to do this. a 300 x 300 pixel image approaches the 640×480 that we see in standard TV, so for on-screen viewing it is pretty good. And this is just the first generation experimental device put together by a bunch of college students… in the long run this is a ‘pivot point’ technology in photography. Imagine being able to watch a documentary on TV and being able to focus your TV on the far mountians or on the stream in the foreground as you wish.
This has a huge potential if you understand the photographic implications.
I have recently been spending a fair amount of time reading the forums at the DP Review site – since I am interested in getting a new camera in 2006.
It has been surprising to read all the complaints about focusing problems, and depth-of-field issues which people are discussing.
With all the arguments over how many AF points are needed, and the DOF advantages of Canon’s “Full frame” sensor versus the APS-sized sensors, there should be a huge number of photographers interested n this technology.