Lytro Photography

[cliveh]

I was fascinated by this concept and perhaps this is fanciful thinking, but wondered if it would ever be possible to develop a spin off programme that could do a similar or limited job to old photographs with only limited planes of focus.

http://www.lytro.com/living-pictures#living-pictures/282?&_suid=47

 

[nhemann]

while I am certaily no expert on the camera, I have been following these (i'm on the pre-order list for the PC version actually) The camera works by somehow capturing the vector field of the light entering the sensors and then being able to refocus anywhere within the "cube" of information it took. Being that an existing photo is already a plane, I don't think that what you want to do would work. But then I barely understand how the camera works itself, who knows? - would be awful cool if it did though. lol

 

[Maris]

I was fascinated by this concept and perhaps this is fanciful thinking, but wondered if it would ever be possible to develop a spin off programme that could do a similar or limited job to old photographs with only limited planes of focus.

http://www.lytro.com/living-pictures#living-pictures/282?&_suid=47

Lytro is another computer based way of making pictures that may (or may not) map things in the real world. It could indeed be possible to synthesise "sharp" images from "out of focus" images but basic information theory (can't distil more from less) indicates that any appearance thus gained would be strongly fictional.
Like all systems that fabricate pictures from data Lytro has nothing to do with photography in its true sense.

 

[nhemann]

Hi Maris, its a well trod path (and admittedly off thread) but I can't help it.

How do you define "photography in its true sense"? I'm here because I am a traditionalist when it comes to photography but neither am I a Luddite (not saying that you are either, btw) Data processing aside - of which I have no clue - if you knew the vector data of all the differential elements of the photo then its a fairly easy jump to the idea that at any given depth from the sensor you would know what virtual focal length it would take to bring that plane into focus - so I am not sure what you mean when you say it may (or may not) map the real world. Short of being manipulated, I would say of course it maps the real world, just as much as one of our silver prints does.

I think its analogous in some ways to a movie camera, instead of taking many pictures that vary with time, this Lytro camera has the ability to take many pictures that vary with distance. I think its a pretty interesting concept. Am I going to sell off all my trad gear? Heck no - but I'd sure like to give this thing a whirl.

Thoughts?
N

 

[Newt_on_Swings]

Whats the difference between this, versus a image captured on a small sensor at a small aperture for deep dof, and using current selective focus post software?

Company sounds like a lot of hype to me.

 

[polyglot]

Lytro has the founder's PhD thesis on their website, which explains the process and is worth a quick read[1]. It relies on an array of microlenses in the camera to produce an image on the sensor that you won't really recognise as the scene in question, and then they apply a bunch of maths to extract the scene information at any particular DOF or focus point.

You can't apply it to an older photograph that wasn't taken with the array of microlenses.

It's also worth noting that the approach is a tradeoff between planar image resolution and depth resolution. The prototype had a microlens size of 12x12 pixels, which means total image resolution was 144x lower than that of the sensor, in exchange for 12 steps of depth information. There's a reason all the demonstration images on their website are only a couple hundred pixels across and still blurry; it's all the technology can manage.

I suspect that application of this microlens technology and processing to a large-format image (probably with a scanning back) could yield enough information to make it worthwhile. Of course, the huge sensor means natural DOF is even smaller, so you probably have a commensurately greater resolution-loss factor to obtain the same depth information... but I ain't sure of that.


[1] in fact, please read and understand it before making uneducated guesses about what they're doing. This really is a new technique that really does work, but it has tradeoffs.

 

[holmburgers]

It's not hubub.. it's a fundamentally different way to take pictures!

If you can't belly up to the bar & accept the development of technology as a positive thing, then you're a luddite.

Like polyglot explains, it's a plenoptic camera and has some analogs to integral imaging with fly's eyes arrays, an idea first proposed by Gabriel Lippmann at the turn of the century (not the more recent one, mind you).

Maris, your assumption that the information is somehow fictionalized and that it may or may not map things in the real world is wrong. This kind of camera technology is an outgrowth of brilliant analog photography experimenters from over 110 years ago and your berating of non "traditional" analog techniques is an afront to their legacy.

 

[cliveh]

The original post was not intended to raise debate about its photographic validity, but merely to promote discussion as to whether this type of technology may one day provide a stepping stone to reveal guesses about out of focus areas in historic chemical based prints.

 

[holmburgers]

Sorry, I think most of the contention was raised by Maris' claim.

The answer is no, it's impossible. There is no guessing involved; the Lytro camera captures the original 3D scene in a fundamentally different way in the first place. Starting from a 2D print, there's no way to regain that infromation that was never captured in the first place. If there was some way to make sense of out of focus information, it would be of an entirely different nature altogether.

 

[cliveh]

Sorry, I think most of the contention was raised by Maris' claim.

The answer is no, it's impossible. There is no guessing involved; the Lytro camera captures the original 3D scene in a fundamentally different way in the first place. Starting from a 2D print, there's no way to regain that infromation that was never captured in the first place. If there was some way to make sense of out of focus information, it would be of an entirely different nature altogether.

50 years ago we would have thought mobile phones impossible. I am not suggesting that the Lytro technology provides the solution. However, if a multiple focus depth of picture is possible, is it not possible that some time in the future an inverse concept of this programme could be possible?

 

[holmburgers]

If you take a snippet of what Maris said, basic information theory (can't distil more from less), this should explain it.

There's no conceivable way of extracting information about the original 3D scene from the antique photograph that now only exists as a 2D object. It's certainly possible to do tricks of photo-editing, but those will be limited by the skill of any operator and will be inherently fictional.

The Lytro literally captures all the information that's necessary to allow focusing later at the initial push of the shutter. It's optical system is wholly unlike a normal lens, and consists of many lenses that allow it to record spatial information, as well as "photic" information (for lack of a better term).

This is hard to explain, but we're discussing totally new stuff, so I guess it's only fair.

 

[MattKing]

It may be the case that there is some unused (and currently unappreciated) information in the 2D photographs that we have now that can be used by new technology either now or some time in the future.

I would think that there might be more potential for this with film based photography, given the fact that a negative or a slide isn't purely 2D - they do have some inherent depth.

 

[Maris]

It's not hubub.. it's a fundamentally different way to take pictures!
Maris, your assumption that the information is somehow fictionalized and that it may or may not map things in the real world is wrong.

Yes, Lytro is a different way of making pictures but not all picture-making processes are photography. It is a general characteristic of pictures fabricated from data, as Lytro does, as digital picture-making does, as painting and drawing do, that the picture-making system is susceptible to data manipulation at the discretion of an operator.

Photography in its original and true sense is one of only half a dozen or so known image making techniques that do not generate or consume data anywhere in the work flow; no data, no possibility of data manipulation. Not only is Photography different to Lytro it is in an entirely separate class of image making.

 

[nhemann]

Photography in its original and true sense is one of only half a dozen or so known image making techniques that do not generate or consume data anywhere in the work flow; no data, no possibility of data manipulation.

Interesting. What other processes do you refer to?

In regards to data/information how do you make the distinction between an array of digital information and an array of chemical information? In both situations the raw data is useless to us until it gets manipulated in an algorthim of some sort. In one case mathmatical computations, in the other chemical process in a prescribed order. Under your definition, doing anything except looking at a scene or, to get past the trivial case, exposing a film and leaving it as a latent image would preclude it from being photography. Using a yellow filter, developing for contrast, burning in a print, even developing at all involves intrepetation, i.e. manipulation of the original data.

Please forgive if you think that I'm attacking in some way as I'm certainly not. I find these grey areas of what is and what is not fascinating to mess around in.

Cheers

 

[polyglot]

Yes, Lytro is a different way of making pictures but not all picture-making processes are photography. It is a general characteristic of pictures fabricated from data, as Lytro does, as digital picture-making does, as painting and drawing do, that the picture-making system is susceptible to data manipulation at the discretion of an operator.

Photography in its original and true sense is one of only half a dozen or so known image making techniques that do not generate or consume data anywhere in the work flow; no data, no possibility of data manipulation. Not only is Photography different to Lytro it is in an entirely separate class of image making.

No offense intended, but that's a bunch of mystical confusion. Photons carry discrete information (finite energy each), exposed/activated silver halides carry discretised (digital) information and the development of film is most certainly the transformation and manipulation of that data. Arguing that digital photography is different from analogue photography is like arguing that drawing with ink is fundamentally different from drawing with pencil because the marker is a liquid instead of an abraded solid. Sure there are (irrelevant) differences in underlying technology but they involve the manipulation of the same information (marks on paper; recording of an optical image) and the user interface is identical (stick makes marks; a planar sensor measures integral of luminous flux wrt time).

While Lytro contains a 2D image sensor, the image on the sensor is not the image as we perceive it because the optical system is not a classic lens. So Lytro is a fundamentally different scene-recording technique from 2D photography but the reasons for that lie in the optics and how they transform the received lightfield into a 2D image, not the fact that there is digital signal processing involved.

 

[Bill Burk]

Interesting. What other processes do you refer to?

I imagine acoustic sound recording would fit the description. Record collectors prize records prior to tape because they are direct recordings of the original sound.

Horn - Diaphragm - Needle - Wax - Stamper - Shellac - Needle - Horn.