Here's an article I think.....

[Bill W]

........everybody will find interesting.


http://techcrunch.com/2011/06/21/lytro-launches-to-transform-photography-with-50m-in-venture-funds-tctv/


Hope you enjoy.


Bill

[btaylor]

Interesting indeed. When you zoom into the sample photo there

[William]

Interesting post!


Hmm, (finger pressed to lips), my <span style="font-size: 9pt; color: black; line-height: 115%; font-family: 'Arial','sans-serif'; mso-fareast-font-family: Calibri; mso-fareast-theme-font: minor-latin; mso-ansi-language: EN-AU; mso-fareast-language: EN-US; mso-bidi-language: AR-SA;"]scepticism is raising a red flag.


" Rather than just capturing one plane of light, it captures the entire light field around a picture, all in one shot taken on a single device. A light field includes every beam of light in every direction at every point in time."


Very bold claims indeed. How do we interpret such a satetment? "Every point in time" - does that mean the photo I take now captures not only the moment the shutter is released but also the image as it appeared five minutes ago or even five minutes in the future?


"entire light field around a picture" - Do they mean around the Subject or around the Picture? They say picture but the implication of the paragraph seems to suggest the Subject, <span style="font-size: 9pt; color: black; line-height: 115%; font-family: 'Arial','sans-serif'; mso-fareast-font-family: Calibri; mso-fareast-theme-font: minor-latin; mso-ansi-language: EN-AU; mso-fareast-language: EN-US; mso-bidi-language: AR-SA;"]especially when they say things like "A light field includes every beam of light in every direction" and "Experimentation... with 100 cameras in one room". Even if they do only mean just the "picture" (from one vantage point) then we would still be dealing with the full dynamic range and full light spectrum, both visible and invisible.


It's like the matrix meets surround sound for the eyeballs.


If every statement in this article is to be taken literally then I envision law enforcement taking a picture of you speeding on your motorcycle, spinning the image around so they can see the rear number plate and doing all this the day before you got on your bike. Amazing!

[neuroanatomist]

Amazing hype, at any rate. Someone posted this on CR as well, so I'll repost my reply from there:


Nice website. I love the clear explanations of the technology, the detailed specifications, etc.

If I had to guess (and some of the limited information in the site supports this), it's primarily software. They've built a camera that's more computer than camera. Take a fairly wide aperture lens, capture a short, fast frame rate video clip as the focus elements are automatically racked from one end to the other (i.e. make a focus stack of images), then the software performs blind deconvolution on the image stack and saves out the deconvoluted cube as a new image stack.

The process has been in use for quite some time in microscopy, with turnkey deconvolution systems available for over a decade. When I used some of the early systems, processing the image cubes would take a several hours on an SGI Octane workstation (turned out it also took several hours on a Cray X1 supercomputer - but the Cray could process a few dozen image stacks simultaneously, vs. just one at a time on the Octane).

[William]

Hi Neuroanatomist, this seems to be your field of expertise. (Asside from the power of the processor) Would you therefore interpret this device to be more about depth of field rather than composition and dynamic range?


The time difference between the first and last frame would have to be very short for moving subjects. I wonder how well it would perform in low light situations given this constraint? Also I

[neuroanatomist]

Would you therefore interpret this device to be more about depth of field rather than composition and dynamic range?


The time difference between the first and last frame would have to be very short for moving subjects. I wonder how well it would perform in low light situations given this constraint? Also I'd imagine the file size to be rather large. How do you see this playing out?

After doing a little more digging (Google search for "plenoptic camera" as a starting place), it's about simultaneously capturing multiple focal planes in a singe image. Early versions used multiple cameras for that, the planned implementation here is a set of microlenses in between the main lens and the sensor - relying on a high pixel density sensor that is, in effect, subdivided into one mini-sensor for each microlens.

[HDNitehawk]

Why not just bring all in to focus at once? Couple this with street view on google earth and you could take a road trip any where you want without leaving your house.


When I first started reading the article I thought it was headed to expanding the dynamic range. When the hype said it would capture all the light.


This is a PR article, they still have to make it happen economicaly. There was a firm here in the state selling advance tickets for tourist space flights. We even have our own space port, but unfortunatly the company wasn

[William]

What am I suposed to do with this ticket then?

[HDNitehawk]

What am I supposed to do with this ticket then?
<div style="CLEAR: both"]</div>

Send it to me. I will put it in a frame with a few hundred shares of ENRON I have and it will make a good conversation piece.

[Kayaker72]

This article reminds me of one I read a long while back. It had quotes from tech types that were literally giddy, made claims that transportation would be forever revolutionized, cited patents for unique gyroscopes/etc, and speculated that the invention, when unveiled, would maybe be something like the vehicles in the Jetson cartoons.....and this was in a major publication.......


Yep...when it was unveiled it wasDean Kamen'sSegway.......