How Can This Be? An Image You Can Focus After You Capture It

Today all the buzz in the photo world is coming from an unusual tech company called Lytro. A new camera called the Lytro Light Field Camera uses a groundbreaking new sensor design that allows more light to be recorded than normal sensor technologies. This in turns actually allows you to choose your focus point after you have taken the photo (click on any of the photos in this post). As of right now no major camera companies have picked up this technology for their own cameras, but worry not as Lytro is working on creating a small point and shoot camera that will feature this "focus later" sensor soon. Check out the example images below as well as some commentary from the Wall Street Journal. Do you guys think this is a good idea or a practical application of selective focusing?

Patrick Hall's picture

Patrick Hall is a founder of Fstoppers.com and a photographer based out of Charleston, South Carolina.

Log in or register to post comments
40 Comments

Pretty cool!

It sounds too good to be true, I call it fake...

This is nuts... Is this another (but late) April fools joke? If this is really -- amazing. I remember a while ago there was news on a "multi-lensed" lens that would essentially take numerous photos focused at different points. Later, combining the images would yield an infinite-depth-of-field image...

What I really want to know is how this works...

And thanks to the fstoppers crew for all of the suggestions for my short time in Charleston. I had a great evening. Hope your wedding shoots went well :-)

I hope it's not an April Fools joke...it was on Time, PCWorld, Wall Street Journal, and tons of other legit sites. 

Where did you wind up going in Charleston?

LOL. You have to be able to take a decent picture to take advantage of this technology. It also won't work for every picture. Also, after the picture leaves the "Lytro" software... and it flattens the image... no more "fun". LOL. Great consumer product if one can manage to take a picture well enough to take advantage of the software. Seems gimmicky.

and the image quality on these photos looks terrible.

nice! shoot wide, don't worry about focus, and you can go back to your computer and crop, adjust focus, and apply other adjustments later. if it works and they provide decent quality, who needs photographers? photo editors is going to be where it's at.

This is not nuts or fake. Look up the scientific papers about it. They really published this!

consumer cam with such bokeh?!

I was wondering the same thing, a point-and-shoot should not produce images that look like that.

If this is not a fake, could it be the other way around than we think it is?

I mean, a small sensor with a small aperture could pretty much render everything in focus. Then the software "adds" the "bokeh" and you'd be able to move the point you want to be sharp and the software blurs everything around it.

That is exactly what I was thinking when I was watching it.

Game Changer!

This will spawn 2nd and 3rd generation technologies until current lenses are rendered obsolete.

On the up side, no doubt it will be less expensive then the current BiG BoY glass that you aces are using now!

Let's see how this all shakes out down the road...

Reminds me of : http://www.popsci.com/gadgets/article/2011-05/cameras-40000-lenses-help-...

I put it in the same category as the hipstamatic iphone app.  Instead of being a good photographer and getting the shot the way you planned using your own creativity and skill, you just play around until you get what you want.  No skill.  I hate it.

Plenoptic imaging isn't new folks. Very real, actually.

Bryan is spot on. Google Plenoptic and you'll see the only "new" thing here is the idea of shrinking it and packaging it for consumer use. From what little I know about plenoptic imaging, I bet the resolution sucks (which is why it is for consumers because really who needs more 3MP for pictures that will just end up on facebook for 5X7's?)

Ok at first I was thinking fake but then again its been all the talk on legit websites and media. However I dont see it being the complete future. There is no way of controlling the Depth of Field which is really a key to great composition. Looks like a nice tool for a smart phone of the future thats about it.

I have read somewhere that it breaks up your sensor into a large amount of smaller sensors and it saves a a copy of the image with a different focus point in each. So you have a nice "focus anywhere" image, but it's minimal in terms of resolution. Your 18 MP sensor, split into 18, gives you only that much.

Not trying to be negative but it sounds like another fake camera like that clear box camera thing you guys had on here a few months ago.

The difference between this and the invisible camera was we sort of broke the invisible camera and hinted at it being fake.  This camera is on tons of legit websites...we aren't pulling one over you this time :)

I am reminded of the clear pinhole camera with the super fast shutter. It's on the internet. it must be true.

i dont know about this camera.

Dr. Ng developed Fourier Slice Photography during his PhD studies at Stanford.  The potential problems in application are:
1. a worse resolution-dynamic range compromise point
2. insignificant selectivity in focus limited by optics

tons of micro lenses capturing multiple images onto a smaller than full-frame sensor? Doesnt sound like the photo quality will be all that great, and the examples are just reinforcing that viewpoint.

I did not read the technical papers now, but I saw something similar a couple of years ago. Not sure if it was a 5d or 1d, but the research team essentially replaced the camera sensor with multi-layered setup, allowing them to measure the light as well as the angle of incidence!
Tracing my history of this I found that I read it at http://graphics.stanford.edu/~renng/ (Redirects us to the new company) and the original wired article at
http://www.wired.com/science/discoveries/news/2005/11/69594

When I read this originally, it prompted me to play around with doing this with a single exposure, resulting in:
http://frostymonkee.deviantart.com/gallery/#/df9y9y

Apologies, technically a dead site :)

This is still not comparable to DSLRs even if you can adjust the point of focus and blur out other parts of the image, your exposure must be set for the entire image leaving parts of the photo over/under exposed. Id say stick with choosing a POF in camera and create art that way, it allows you to put more thought into the image.

It's not very precise... and would this be an advantage, or a detriment to creative photography?

I am really curious as to how fast the lenses are going to be on these cameras. If it is at all related to the physics of current lenses, I would think they would be terrible low light cameras as they are going to have pin-hole apertures in order to capture all of the scene detail for later manipulation through the software. Pretty interesting technology. I can see the consumer side uses being far greater than the professional side though. 

Like... Blade Runner

The examples they show could easily be faked, i made a quick example last night.
http://birdy58033.webs.com/Focus%20Slider.html
Took the HDR approach, a bunch of images then squished them together with some quick javascript.
Mine are obviously different images, as I shot a windy tree. But really all you need is a good fps.

where is your tripod JW?  Of course this could be fake...but it could also be real :)

It could totally be real! But, Lytro plans on putting out a point n shoot with this tech. And, they didn't show any actual hardware. Which means they may just be modifying a high mega-pixel camera to shoot several images really quick then stack them in software. And then give the camera itself a lower megapixel rating. For a pro photographer this won't work, which may be why they are targeting the point n shoot crowd. When a big company like Adobe says the tech is not currently feasible for consumers, and a small start-up thinks they can make it cheap enough for the point n shoot crowd, there must be some sort of compromise going on.

Adobe did a presentation years ago in a Photoshop demonstration of things they were testing for future versions. Manipulating images taken with a plenoptic camera was one of them. It is real, the tech has been here for a while, but like others have mentioned it cuts down big time on the resolution. If they can figure out a better way to do it then it will be more photographer friendly. Those that criticize the quality of the images can take that up with the person who took them. As far as resolution of images that is a decent concern, then again I doubt this was aimed for the pro market, who should be able to focus their images properly the first time. What would be cool is a plenoptic video camera. Imagine filming without the need to do manual focus pulls. :)

Um focus stacking? I believe I learned that here. Anyways, make a product, get people to believe in it, and create a press release that you send to multiple news outlets and bam, you're gonna get noticed.
Oh, what's the deal with the "black" images on their websites in the first place? At lease the invisible camera was interesting...

adobe first showed the ability to do this a few years ago except they were a little more extreme being able to paint focus, and being able to change the the angle slightly.  they used a 28 lens array to achieve this it took 24 hours to render.

I'm not surprised or doubtful at all. Just wait until the lighting software takes over. We'll be able to manipulate light in post using virtual soft boxes, windows, sun etc. Cameras will record distance points giving a 3D type image for software modeling of any kind of light. Focus, f-stops, speed and lighting will all be done in post.

Wait! Their trying to make a camera? Where's the camera that took these pictures in the first place? And some of the photos on the site have like 3 focus points, there was one where I click on the roof in the background and it focused on the closest thing next it which was a flower that in reality was about 10 feet away from the roof. FAKE

Mankind seems obsessed with developing the easy route for the masses, and offering situations with instant satisfaction. Why bother learning, putting effort in and even thinking.

TechCrunch interviewed Lytro CEO and founder Ren Ng. Quite an impressive session with some details about its immersive 3D feature. Check this out:

http://techcrunch.com/2011/06/21/lytro-launches-to-transform-photography...

and the best part? this camera and its film are TOTALLY INVISIBLE! hahahah

Now.

Combine This sensor and the Dragon Sensor.That now imagine what shots you could get without a focus pull.