Imaging Science Breakthrough of sly scam?

[Michael West]

lytro...claims they have develope an entirely new type of sensors that can capture images which can be focused...after the fact.

I sense they arent being particularly forthright with this but know so little physics that I cant say much with any degree of certainty

I was under the impression that LENSES rather than sensors were what focused images.


http://www.lytro.com/

[Christoph C. Feldhaim]

The additional recording of the angle of a light ray allows these special features, but it comes at a price: resolution.
The dissertation of their CEO available at their website is an interesting read - you can skip the all too complicated mathematical stuff when reading.
I think we'll have to wait for the real world product to see how it really works, and I believe it will take some time until it makes its way to high resolution photography.

[feppe]

There have been tech demos of this for years, it's the real thing: there are several academic papers on this, and here's an article from 2008. Whether the product is the real thing is another matter.

[Michael West]

The additional recording of the angle of a light ray allows these special features, but it comes at a price: resolution.
The dissertation of their CEO available at their website is an interesting read - you can skip the all too complicated mathematical stuff when reading.
I think we'll have to wait for the real world product to see how it really works, and I believe it will take some time until it makes its way to high resolution photography.

The current 3 megapixel ceiling seems a bit limiting, and points toward a long road before its ready for "prime time"

Thanks

[feppe]

The current 3 megapixel ceiling seems a bit limiting, and points toward a long road before its ready for "prime time"

3 MP is plenty for screen and web use, and the tech needs a computer and screen to be fully utilized. Printing photos is already a niche activity, anyway.

Besides, it was only a few years ago 3 MP was a lot in a compact camera, and they are now in the 12+ MP realm.

[PierreVandevenne]

This may be an area where very small pixel sensors could be useful in the future (especially since the principle of the images combination/practical downsampling itself leads to noise reduction). And of course, it will depend a lot on the micro-lens array characteristics. If the idea catches, we'll probably see a lot of progress in terms of suitable micro lens array manufacturing.

AFAIC

Coolest practical thing in the paper: the camera could have a dual personality. Forget flexible focus and get full resolution, or close. Sacrifice resolution and get flexible focus and even a certain amount of 3D imaging.

Most amusing idea: working with current sensors, a camera with a square aperture should be better at the task.

Quite fun to imagine, a few years from now, a camera that would automatically operate optimally from a DR point of view (forget ISO), would seamleslly switch from max res to max focus depending on the subject and that would have a... square aperture.