Previous news story    Next news story

Just Posted: Lytro Light Field Camera review and video

By dpreview staff on Feb 29, 2012 at 22:33 GMT

Lytro has shipped its first Light Field Camera to a customer and we've had a chance to spend some time with one, to see what their experience is likely to be like. It's a totally unconventional camera that captures images that can be refocused after they're shot, so we haven't shot our usual, 2D test charts but we've tried to sum-up its technology and what it's like to shoot with.

Review video

8
I own it
9
I want it
1
I had it
Discuss in the forums
10
I own it
17
I want it
1
I had it
Discuss in the forums

Comments

Total comments: 310
12
djbrom
By djbrom (Mar 1, 2012)

Interesting but it doesnt really grab me.

More interestingly and kinda off topic..... What did you use to record the video? :)

Comment edited 20 seconds after posting
0 upvotes
Le Kilt
By Le Kilt (Mar 1, 2012)

and the reviewer should turn to the camera that is filming, i.e. no good using three cameras if he only speaks to one - but I'm sure next video will correct this :-).

0 upvotes
ProfHankD
By ProfHankD (Mar 1, 2012)

The Lytro offers very clever, sexy, Apple-inspired packaging as the lead-in to a variety of computational photography processing methods. The real question is: Do people find after-the-fact playing with focus compelling? I think stereo capture is more compelling, but refocus has no competing products, whereas stereo capture has competitors using pairs of lenses without computational photography processing.

Plenoptics are not the only means to the same set of computational photography goals; for example, similar things can be done with coded apertures. I've been developing computational photography methods using a sub-$1 color-coded aperture that allows single lens, single shot, anaglyph capture with commodity still/video cameras -- and reprocessing to stereo pairs, refocus, etc. In short, there's lots more to come as digital cameras start to more aggressively leverage being computing devices... Lytro may play a big role in making the public understand this.

1 upvote
Felix E Klee
By Felix E Klee (Mar 1, 2012)

"color-coded aperture", how does that work? Any good introduction somewhere?

2 upvotes
ProfHankD
By ProfHankD (Mar 1, 2012)

My Instructable on single-lens, single-shot, anaglyph capture is:

http://www.instructables.com/id/Use-Your-Camera-To-Capture-3D-Anaglyphs/

The slides from my IS&T/SPIE Electronic Imaging talk give a decent overview of computational reprocessing of anaglyphs:

http://aggregate.org/DIT/SPIEEI2012/spieei2012slides.pdf

Obviously, I've been making all of this freely available.... I'm a professor at the University of Kentucky, not a company. ;)

2 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Surprised the University Kentucy doesn't want to control the technology--after all Stanford owns the underlying tech for Google, I'd imagine Lytro too.

Also your example looks like a version of 3D, that's not what this Lytro is doing.

0 upvotes
falconeyes
By falconeyes (Mar 1, 2012)

You're doing great work, thanks for the link. I esp. like the progress done on processing the color information into a relatively good rgb image.

However, whatever be the method (anaglyph capture, plenoptics) you can't escape the fundamental laws of quantum mechanics, esp. Heisenberg uncertainty relationship. By trying to encode/capture directional information you effectively lower the aperture (directional uncertainty, momentum) for photons when detected with a given positional uncertainty (pixel size). Anaglyphs use color coded small apertures and plenoptics (Lytro) uses a microlens array to the same purpose.

You always end up with the same formula: the depth of the refocussable region "times" number of pixels in focus is a constrained number for a given size of lens. I oversimplified the formula but you get the idea. You have a relatively narrow refocussable region (you still focus your camera) with fair resolution, Lytro choose a deep refocussable region with poor resolution.

0 upvotes
ProfHankD
By ProfHankD (Mar 1, 2012)

UK does like to control these things, and I did the appropriate IP disclosure, but UK didn't follow through with a patent and I'm a tenured full professor with an endowed chair, so I can get away with doing good things for the community. Also, recall that unlike Stanford, UK is a public institution where doing good for the community is an official priority.

As for 3D being different from what Lytro is doing, well, no. Both are just leveraging the fact that different viewpoints pass through a single lens and can be separated. The methods for separation are different, and they did refocus first while I've done stereo first, but the two technologies are largely interchangeable, with various tradeoffs on processing complexity, image qualities, etc.

1 upvote
Felix E Klee
By Felix E Klee (Mar 1, 2012)

Wow - I have to try this out some day!

If I understand your approach correctly, then you get two rays per dot, while Lytro gets quite some more, and in full color.

Have you every tried more rays than two, for example by taking a series of pictures and moving the hole in front of the lens? Would that work? And would that allow refocusing later?

1 upvote
ProfHankD
By ProfHankD (Mar 1, 2012)

Normal anaglyph viewing and my default processing does precisely that: two full-color half-resolution views are generated.

There is no need for more or moving holes. Computational reprocessing treating the anaglyph as a coded aperture image can resolve much finer view angles without loss of resolution. The catch is that the color and view recovery is a computationally much harder and less reliable process than one gets with the Lytro. The Lytro has other advantages by not using a commodity camera -- e.g., the Lytro's lens was probably tuned to have far less artificial vignetting than most commodity camera lenses, and this property is critical for both approaches.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

ProfHankD:

Yes, UK is public, but Stanford gets a fortune in public monies for various research projects.

0 upvotes
CyberAngel
By CyberAngel (Mar 1, 2012)

They really need a huge Nokia 808 -like gigachip and an ordinary camera frame with HD screen plus application on a Mac/Win/Linux
add 3D shooting with very fast focusing and then the money comes
Maybe Sony could bring it on?
I think that even Lytro themselves has missed the point of light field technology:
It's for 3D
Movie camera could be the best ever product

0 upvotes
Bill1969
By Bill1969 (Mar 1, 2012)

One useful thing is to take a shot and process it to generate a JPG version with every object in it focused, this is useful for landscape or wide angle photos.

Maybe when the technology develope better, if the camera has fast shuetter speed, it would Be nice to take sport or wildlife shots without worrying about focus..

Comment edited 2 times, last edit 8 minutes after posting
0 upvotes
realityChecker
By realityChecker (Mar 1, 2012)

...Part 2

So beware technology hypers', you just go ahead and tout all you want about the futuristic line like.... "just imagine digital photography 20 years ago, blah blah..."....but here, the fundamentals are different. This is a giant lens projecting on to a VERY small sensor as far as I can see, (if you know more please tell me, I'd like to know) but this to me seems like shooting f128 all the time, so OF COURSE everything is in focus. Now, if you took 500 lbs of glass and shot it at a 12MP, I'm sure you could get everything in focus too on that sensor as well. But I'm only interested in shooting at f 2.0 at 12fps so you won't get my dollars. Otherwise, enjoy! :))

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

This is an F/2.0 lens--only ever F/2.0.

You neglected the prisms in the system.

0 upvotes
kdaphoto
By kdaphoto (Mar 1, 2012)

Why are you so threatened by this product/technology?

1 upvote
realityChecker
By realityChecker (Mar 2, 2012)

@HowaboutRAW: f2.0 relative to what sensor size?...do I really neglect the facts?...perhaps not as much as you tout the fact that an "after-the-fact" Gaussian blur will dramatically change photography as we know-it. IMO, this "blur" looks like garbage and the number of prisms you have to me don't matter for hand-held photography because one can achieve that "in-focus" effect without them. but then again what for?...an all in-focus shot that even a decent POS can take now with enough light?

0 upvotes
realityChecker
By realityChecker (Mar 2, 2012)

@kdaphoto: look, I'm not threatened by this technology. I'm threatened by people who propagate solutions to problems that don't exist and I feel that investors should put their money into something that will really make a difference in the world. If this technology works in some niche markets, I'm very happy for you. But to tell me that it replaces optical-based bokeh in creative photography and I'll be firm with my opinion about how I feel about that. I'll say it's garbage and not because I'm threatened, because I think it's ugly and IMO useless because I like my shots in... and out of focus. :)

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 2, 2012)

realityChecker:

That's F/2.0 relative to the whole sensor.

You really need to make some effort understand this technology before spouting.

0 upvotes
realityChecker
By realityChecker (Mar 1, 2012)

Part 1:

Lets try to and begin to uncover the story behind the hoax shall we?

Let's just say we take a HUGE lens and mount a tiny sensor behind it which is SO small that EVERYTHING is in focus. Now, let's call-it a "technology" and sell shares to gullible investors that think this will evolve.

Then sell a few and give away more and get the consumer to log-in and choose a "focus point". Then use some generic Gaussian blur algorithm to blur the rest of the photo in a way that only a fool who's a totally inexperienced photog thinks that it's bokeh (when it's really) NOT. From what I see, this looks and feels NOTHING like true OPTICAL bokeh and I don't see that there is a way that it could ever simulate true lens bokeh.

....(con't)

1 upvote
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Not at all how light field pictures work, try reading up on it first next time.

1 upvote
Cy Cheze
By Cy Cheze (Mar 1, 2012)

Why is a "plenoptic" approach better than a deep focus photo that one can edit to defocus selective portions?

How will any part of a "plenoptic" photo be in focus if the shot is hand-held, the light is bad, and the shutter is slow?

0 upvotes
kdaphoto
By kdaphoto (Mar 1, 2012)

Your sarcasm betrays your ignorance on the technology, or the lack of desire to understand it.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Cy Cheze:

Shutter speed doesn't control focus.

0 upvotes
realityChecker
By realityChecker (Mar 2, 2012)

@HowAboutRAW - Although my sarcasm is perhaps not nearly as ignorant of a statement than Lytro's CEO who said that "Regular photographs just don't tell the whole story." So please, spare me the drama. Then the PC mag article(linked below) give us a terrible photo with models blurred in-behind the main model. That to me is simply a crappy photograph in my opinion, (but then again, I guess you'll say it's subjective and you'd be right?). So I'm just stating that I still don't think people will buy this PR stunt, prisms or magic potions. I'll gladly keep the traditional bokeh from a chunk of glass moving glass closer to the sensor thank you very much. Now if you can come-up with an algorithm that will truly simulate that, I'd be impressed, but till then, I'm sure more PR people will come-by and "nay-say" negative posts like mine and these PR people will try to confuse people in this spin zone. To that I say enjoy the spin zone! ;)

http://www.pcmag.com/article2/0,2817,2387554,00.asp

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 2, 2012)

realityChecker:

Still have no idea what you're talking about.

You appear be confusing Lytro's execution in 2011/2012 with what will likely be a much improved application of the optical principles in 20 years.

You still appear to have made no effort to understand how this gear works.

Saying there are flaws in Lytro's particular example is next to meaningless except as an objection the this particular light field camera--not those cameras generally.

Comment edited 6 minutes after posting
0 upvotes
jkokich
By jkokich (Mar 1, 2012)

I don't get it.

0 upvotes
Russell Dawkins
By Russell Dawkins (Mar 1, 2012)

I mean it literally - I expect we will see a forum for Lytro users in due course, if the camera takes off - won't we?

0 upvotes
Russell Dawkins
By Russell Dawkins (Mar 1, 2012)

I wonder when we will see the Lytro forum...

0 upvotes
CameraLabTester
By CameraLabTester (Mar 1, 2012)

Okay lets cut to the chase...
Could anyone please list down the10 great things you can DO with this LFC:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Thank you :)

.

2 upvotes
Lillrobert
By Lillrobert (Mar 1, 2012)

It could be verry useful for macro shooting if it had the right lenses and a proper camerabody with lensmount and if you could render infinitive DOF image to get the whole object in focus.

1 upvote
Najinsky
By Najinsky (Mar 1, 2012)

1. Nudes - always fun to interact with
2. Photograph ethical News International staff. Has the Resolution.
3. Repeat 2 for Olympus Board
4. Disguised rocket launcher for terrorist posing as photographer
5. Put an LED in each corner for a great fashion torch
6. Replace the lens with a mirror for a portable make-up mirror
7. Prise out the LCD to 90° for around the corner shooting
8. Remove the LCD completely for a novelty cookie cutter
9. Spare table leg
10. Review it and start a photography blog

8 upvotes
kdaphoto
By kdaphoto (Mar 1, 2012)

Why do you insist on making new technology have all the answers NOW? That's pretty narrow-minded thinking. How about we let the early adopters play with it and see what happens?

0 upvotes
CameraLabTester
By CameraLabTester (Mar 1, 2012)

I better be carefull with that lying around my desk...

I might poke the lens thinking it was my pencil sharpener.

.

2 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

The whole box is a good bit smaller than those powered desktop pencil sharpeners. More likely to push it off the desk if you mistook the near inch diameter lens for a pencil hole.

0 upvotes
kdaphoto
By kdaphoto (Mar 1, 2012)

What I find interesting is the over analysis of a product that, I think, has been created to introduce an interesting technology to the public and have fun with. It seems that some commenters are putting the expectations of a fully developed DSLR on a brand new type of camera. I think it is a product for early adopters and tinkerers (me included since I've ordered one). Lytro may and should have an idea of what's next for them but I'm sure they are watching what users do with it. I think the Polaroid reference is appropriate because the Lytro is something to go out shoot, experiment, have fun, and learn how to use what it does. Is it limited, yes, is it perfect, no but so what.

2 upvotes
the great fiction
By the great fiction (Mar 1, 2012)

I think you're spot on. We would all be worst off if innovators waited to bring their products to market until either 1) it pleased the users of the existing technology (in this case, the traditional digital camera, whether compact or DSLR) or 2) they had figured out exactly how users would use their product.

2 upvotes
the great fiction
By the great fiction (Mar 1, 2012)

In a similar vein, you have to think about this in terms of "disruptive innovations," i.e. new products that target a neglected market and appear to not threaten the dominant players (e.g., Sony's transistor radios weren't a threat early on to RCA's tube radios because the sound quality was inferior and mostly only teenagers bought them). But once the technology gets far enough along and can compete with the dominant technology, it's too late for the companies that rely on the old technology to adjust and adopt the new technology.

2 upvotes
kdaphoto
By kdaphoto (Mar 1, 2012)

Based on many peoples reactions, it seems that this product/technology somehow threatens them. How about sitting back, experimenting, and having FUN. Aren't we supposed to do that every once in a while? My delivery date isn't until April/May, I'm already impatient to get my hands on it and just play.

0 upvotes
SheikYerbouti
By SheikYerbouti (Mar 1, 2012)

This is very impressive technology, indeed. But I still find it hard to get my head around the underlying working principal of this camera. No doubt, it would be fun to toy and experiment with it for a while. But in the long run? I'm not sure, even if the resolution was adequate for quality photography. It seems to go against everything I have ever learned about photography.

Another issue that no one has addressed so far, is that this technology seems ideally suited for the spy/kill drones of the future. And this is an aspect I don't like at all. Why does stuff like this always have to end up in the wrong hands?

1 upvote
Nightwings
By Nightwings (Mar 1, 2012)

The wrong hands? You mean like the hands that try to stay one step ahead of those that would do us harm?

0 upvotes
SheikYerbouti
By SheikYerbouti (Mar 1, 2012)

You won't see it that way, when your beloved government no longer only uses such technology against kindergartens and wedding parties in far away countries but also against their own domestic riff-raff, as they most certainly will at some point in the near future.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Nightwings:

Because those wedding parties are out to do "us" harm. (I'm being ironic, to be clear to those who'd agree with my statement.)

Comment edited 1 minute after posting
0 upvotes
falconeyes
By falconeyes (Mar 1, 2012)

R. Butler:

You wonder what a 41MP sensor would give. Well, with the current hardware (pixel size and F-Stop), it wouldn't help. Reason: the very idea of capturing light fields meets the tight constraints set by the Heisenberg uncertainty relation and cannot be avoided. Assume that 3x3 rays per pixel are recorded (which is in line with a 1080^2 resolution and 11 MRays), then each ray is recorded with the accuracy of F6 (3x F2). The diffraction limit for F6 is 2µm, the pixel size (using the data from FCC) is about 1.5µm. This shows you that Lytro already exploits the very limits of physics. A higher rez sensor wouldn't help.

In order to obtain higher rez, you need more image pixels and more rays per image pixel to keep the hyperfocal distance. This makes the sensor size grow rapidly with resolution, 2x the resolution probably means 16x the cost, needing a large 160MP sensor ... The 1.4µm 41MP sensor probably would only allow 1300^2 resolution.

0 upvotes
infosky
By infosky (Mar 2, 2012)

This is a great insight. But, after I analyzed the problem, I think your argument might not be completely accurate. The diffraction limit of F6 actually takes place at the front focal plane of the micro-lens, not the sub-pixels. The micro-lens performs another Fourier transform. The F/# of the micro-lens determines the diffraction limit of sub-pixel.

On the other hand, I think there is an inherent limitation of the sharpness (MTF) owing to the discreteness of microlens and sub-pixels and the loss of the phase information once the light is detected. Without careful consideration, there are potentials for cross talk between pixels.

0 upvotes
infosky
By infosky (Mar 1, 2012)

If I go to a party and take 100 pictures, I will have to refocus everything picture before I choose a few to share with my friends. Who will do that? Who has the patience to upload 100x10MB data to Lytro website before seeing the first picture?

I think the concept of focusing after shooting is completely flawed for casual users and it is even more flawed for serious users.

So, even the resolution is not an issue, it will not be user friendly unless the post processing is no-brainer and nearly instant.

Finally, for those who post pictures on Facebook, the quality of picture is not important. Any P&S camera is more than enough.

3 upvotes
sedentary_male
By sedentary_male (Mar 1, 2012)

Wait for the smartphone app. Then what happens to the Lytro?

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

sedentary_male:

The smart phone would also need a prism inside. So an application for smartphones isn't going to suffice.

0 upvotes
sedentary_male
By sedentary_male (Mar 1, 2012)

I'm struggling to see the point. Who is this aimed at - really. Would a customer that spends US$399 or US$499 really be bothered enough or even able to open up the files on a PC and adjust the focusing?? Or, are the type of photographers who would do that be satisfied with the form factor , interface , feature set etc..

1 upvote
Ben Tomohiku
By Ben Tomohiku (Mar 1, 2012)

As I understand, they are trying to make photography even easier to people, I can't agree with that. It sounds "Fun", but not for me personally. I really enjoy the manual camera, because I feel taking control by using my knowledge from years experience. Maybe that crazy fully auto camera takes wonderful image, I would be happier to see my imperfect one taken by "Me".
Lytro introduced a really nice concept, however I can see the day that Canon, Sony or Nikon etc purchase its technology and apply to more mature DSLRs.

0 upvotes
DioCanon
By DioCanon (Mar 1, 2012)

HOW ORIGINAL!
they said the same about digital...
the digital camera that now you are so proud of...
(-too easy you can see already the image on the display...)

Comment edited 45 seconds after posting
2 upvotes
Ben Tomohiku
By Ben Tomohiku (Mar 2, 2012)

I actually shoot film as well, maybe guys like me are old fashion.

0 upvotes
DioCanon
By DioCanon (Mar 1, 2012)

the concept is fascinating, no doubt about that.
but the camera has to be improved in every aspect.
said that I remember my first camera phone, Moto V600 and I see phones now! and I also remember old digital cameras...

2 upvotes
MP Burke
By MP Burke (Mar 1, 2012)

It;s too early to say whether this will catch on, but, if I'm reading it correctly, a 16MB file is being used to create images of about 1.2 MP.
Thus it seems likely that if the resolution was increased to around 10MP, the file size would exceed 100MB. This alone may be a sufficient reason to restrict the application of this type of camera.

Comment edited 6 minutes after posting
2 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

And Samsung NX200 raw files are already 45MB--barely more than a factor of 2 off what you suggest.

0 upvotes
Ben Tomohiku
By Ben Tomohiku (Mar 1, 2012)

It spends too much space to record "light's" information in order to change the focus point I guess. I would rather focusing carefully on what I want on my 5D2, instead of take a cheesy light field photo and try to change the focus points.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Ben Tomohiku:

But you can't shoot every 5D Mark II shot at F/2.0. Also the 5D is the result of many generations of consumer DSLR development.

0 upvotes
love_them_all
By love_them_all (Mar 1, 2012)

If u are serious enough to shoot with a 5d2, you should at least learn how to pick your point of focus. In 2012 tech we are not at a point where pj can shoot basketball games at hi res and no need to focus ala lytro. Too much needs to be compromised.

0 upvotes
Octane
By Octane (Mar 1, 2012)

The effect is cool, the resolution is the deal breaker. It's already very low, but the quality/detail is really poor on top of it. The edges between OOF objects and in focus objects is also not as smooth as you would see it from a conventional lens.

Second deal breaker is that all images only go through their web server. WTH?

As many have said already, it's fun to play with it for a minute, then you get bored because there is no purpose. Refocusing a photo isn't anything exciting, there is zero entertainment value.

The simple and cheap filters inside the Instagram/Hipstamatic app offer more creative looks and more fun (thus user satisfaction) even though they are very basic and sloppy. It's irrelevant how sophisticated this camera captures light rays, the end result isn't convincing in terms of quality nor fun or creative satisfaction.

4 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Well there's a real advantage in shooting all photos with a lens all the way open to say F/2.0, you can shoot with a faster shutter speed. And then with this technology you can set the focus point later.

Big deal, this would be like buying a digi cam in 1982, yes that far back.

Also: no the images don't just go thru Lytro's web server--you can produce those images on a Mac, Windows software will follow.

0 upvotes
Richard Butler
By Richard Butler (Mar 1, 2012)

Except you can only then share those images on your Mac or PC by outputting JPEGs. For the image to be interactive, it has to go through their servers.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

R Butler:

Um, I wasn't thinking of sharing. Also can't one simply download the image file to more than one Mac with the Lytro software installed on it. (Nothing in the demo I witnessed precluded that possibility.) And is there anything that then prevents emailing these files as attachments to others with Macs with the Lytro software. Are you saying that Lytro disguises the file and file folder? Or does Lytro make it impossible to get more than one copy of the software for playing the images?

Comment edited 5 minutes after posting
0 upvotes
Krilnon
By Krilnon (Mar 1, 2012)

There's also nothing stopping you from re-hosting the LFP files on another site. There are at least 3 people working on open source LFP viewers, or at least the version of LFP that the Lytro website display.

I'm working on one, for example, and it's really easy to extract the 5-or-so JPEGs from the LFP. It's a little harder to figure out what the depths are, but only because Lytro hasn't yet published any documentation about it.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Krilnon:

Interesting, I was never disputing that one way of sharing was on a webserver.

Good that someone is trying to move beyond Lytro only servers.

Comment edited 2 minutes after posting
0 upvotes
tongki
By tongki (Mar 1, 2012)

flash based application ?

come on, now is the era of iPad,
we don't use flash anymore, that's so yeterday

0 upvotes
Richard Butler
By Richard Butler (Mar 1, 2012)

There's also an HTML5 version that kicks in if it detects you aren't using Flash.

2 upvotes
DioCanon
By DioCanon (Mar 1, 2012)

apple boys maybe!

now the galaxy tablet has got flash!

maybe also apple one day

1 upvote
Brett St Pierre
By Brett St Pierre (Mar 1, 2012)

Flash is history and it's a battery killer. I understand why Apple doens't want it near any of their mobile products. Avoid. Adobe knows it's history too. I think we know enough already that HTML5 is allowing us to move on from Flash.

1 upvote
Felix E Klee
By Felix E Klee (Mar 1, 2012)

No HTML5 version kicking in over here. Always have to enable flash to view the "Living Pictures". Cumbersome.

0 upvotes
Biowizard
By Biowizard (Mar 1, 2012)

Steve Jobs has passed away; iPad 3 being announced within the next 6 days ... are you SURE there won't be Flash? If nothing else, to allow Apple to compete with Samsung?!

0 upvotes
tkbslc
By tkbslc (Mar 1, 2012)

right now this is mainly a toy, but who knows where it will be in 10-15 years. I remember back in 1999, the $1000 1.3MP digital cameras really weren't useful for anything but happy snaps, but now look at them. If you could do this on a large sensor and shoot f2.8 from infinity to mfd and adjust the best DOF and focus like we do any of the standard RAW settings now.

Comment edited 23 seconds after posting
0 upvotes
idbar
By idbar (Mar 1, 2012)

Oh man! I remember I paid $500 for a 2Mpx camera circa 2001, which would consume the batteries in 30 minutes. It was worth not paying for processing though ;-)

Comment edited 1 minute after posting
1 upvote
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

idbar:

And I paid about $800 at the very end of 2001 for a Canon G2, funny thing I still like what it does. And the batteries last. I mostly like the G2 more than the G6. Both have an F/2.0 lens unlike the G12, irony.

1 upvote
trungthu
By trungthu (Mar 1, 2012)

In October, 2002, I have used Fuji FinePix S2Pro (6,1Mp - $2450), it's not a toy.

0 upvotes
Delacosta
By Delacosta (Mar 1, 2012)

But hardly money well spent either.....

0 upvotes
trungthu
By trungthu (Mar 1, 2012)

As I see in the second photo in
http://www.dpreview.com/reviews/lytro/3
When I click to the glass, the leg behind it also be in-focus, and when I click to the guitarist, the leg behind the glass turn out-of focus.
I think, this is only a software game, by separating each plan by software to put it to in or out of focus with some faults. It's only a game, not technology, as they say.
I think so. I hope I'm wrong.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

In fact you be wrong. The camera's sensor captures much more focus information than a normal digital camera.

(The above is in no way a full summary of the technology. Please feel free to read up on this kind of photography, it's not exactly a new invention or idea.)

0 upvotes
trungthu
By trungthu (Mar 1, 2012)

Thank. But if you see in the photo?
Because they use a very small sensor, so all of the field be in focus, after that, they use a software to separate each plan (by use the interpolation), but with something with "opacity", the software can not distinguish it.
I hope they can do more than that.
Thanks.

Comment edited 2 times, last edit 5 minutes after posting
0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

trungthu:

No that's not how it works, the prism immediately in front of the sensor redirects all sorts of extra focus information to the sensor. The software can then sort this information.

You have no idea what you're talking about, please read up on the specific technology and the optical principles behind it.

0 upvotes
trungthu
By trungthu (Mar 2, 2012)

Thanks HowaboutRAW.
I think I know less than you, but please explain the phenomenon that I have seen in the photo in the link I have talk to?
""" http://www.dpreview.com/reviews/lytro/3
When I click to the glass, the leg behind it also be in-focus, and when I click to the guitarist, the leg behind the glass turn out-of focus.
I think, this is only a software game, by separating each plan by software to put it to in or out of focus with some faults. It's only a game, not technology, as they say."""

1 upvote
trungthu
By trungthu (Mar 1, 2012)

Very impressive.
If we can do it in larger format camera.
Thanks.

0 upvotes
whtchocla7e
By whtchocla7e (Mar 1, 2012)

I had a lot of fun playing with the sample image for 5 seconds.
Would I do it again? Probably not.

3 upvotes
fmian
By fmian (Mar 1, 2012)

I'd really like to see someone attach a few of these to a roomba with motorised height adjustment, add a compositional algorithm, and have them replace your average cookie cutter wedding photographer. Only it would be slightly more useful as it would clean the floors as it went by.
Yeah I know, ouch.

1 upvote
Biowizard
By Biowizard (Feb 29, 2012)

The big question I ask myself is, even if someone gave me one of these cameras, would I use it? The answer is yes, probably, for a day or two, until the novelty wore off and I became frustrated with the limited resolution, and need to run proprietary software and/or use web services to render my images.

If I want the effect of a multiple-focus image, I'd far prefer to take a series of stills from a tripod-mounted SLR at various focal settings, and combine these into an industry-standard interactive FLASH file.

Interesting as it is technologically ...

... THE LYTRO IS A SOLUTION LOOKING FOR A PROBLEM.

Brian

5 upvotes
idbar
By idbar (Mar 1, 2012)

"If I want the effect of a multiple-focus image, I'd far prefer to take a series of stills from a tripod-mounted SLR at various focal settings"

Which I guess is a task this camera is aiming at simplifying. Like...
"If I wanted the effect of a moving picture, i'd rather take many many many pictures and put them together...", oh wait, there ARE video cameras now. And they used to have less resolution!

The fact that you're not the target market, doesn't mean there's no target market.

Comment edited 41 seconds after posting
3 upvotes
chlamchowder
By chlamchowder (Mar 1, 2012)

I disagree - the problem is focus misses and out of focus images. This aims to solve that problem by allowing the image to be refocused in post processing.

Currently, the technology has major problems. Hopefully, those can be solved, because correcting focus in post production has a lot of potential.

1 upvote
marike6
By marike6 (Feb 29, 2012)

So the images it produces, like the samples on the Lytro website, are essentially mini Flash applications, where the viewer can click to change the focal point? Is that the gist of it, the "Living Picture", as they call them? Or are there other uses for this camera?

0 upvotes
Nathaniel George Weir
By Nathaniel George Weir (Feb 29, 2012)

Why do they market this camera as a refocusable camera? Why not a camera that has unlimited DOF. I guess consumers like refocusable make it appeal to a different consumer audience.

0 upvotes
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

They plan to add "all in focus" according to someone from Lytro with whom I spoke at a demonstration.

I think that's what you seek.

0 upvotes
Richard Butler
By Richard Butler (Mar 1, 2012)

It's also mentioned on page 3 of the preview.

0 upvotes
RoccoGalatioto
By RoccoGalatioto (Feb 29, 2012)

Who needs it?

3 upvotes
HowaboutRAW
By HowaboutRAW (Feb 29, 2012)

Who needs a photocopier or a digital camera, or a personal computer, or a car phone. Those are all questions that were asked in the past.

I'd imagine you can add more questions to the above list. Like who needs an indoor toilet.

Comment edited 29 seconds after posting
5 upvotes
idbar
By idbar (Mar 1, 2012)

OMG! Who needs an indoor toilet!? All the houses will smell like poop from now on!! Who'll pay for such thing?

Oh wait... kidding ;)

Comment edited 29 seconds after posting
1 upvote
kenw
By kenw (Feb 29, 2012)

So after downloading a flash update, closing my browser, installing it, and coming back I can finally see the images...

Anyway, nice review DPR. I think you about hit the nail on the head in a fair and balanced manner. No surprises really, did just what everyone thought with the limitations everyone expected as well.

3 upvotes
Joesiv
By Joesiv (Feb 29, 2012)

I think the true potential of this technology will come when/if they introduce an API to process the RAW files.

With such an API, not only could we do what they currently do, but we likely would be able to exagerate the out of focus areas, mimicing effects of large aperture/large sensor cameras. We could also see 3d images among other things.

Time will tell. I'd hope their next model was higher resolution with better manual control. The technology has some potential...

1 upvote
Felix E Klee
By Felix E Klee (Feb 29, 2012)

They should completely open the file format's specifications.

1 upvote
PhotonTrapper
By PhotonTrapper (Mar 1, 2012)

I agree with you that there might be some potential (I am thinking for security camera) but how would you construct a 3D image from data collected through a single lens?
Correct me if I am wrong but the way I understand it, the slight incoming rays angle difference this system is playing with at pixel level (to basically register subject-lens distance information) has nothing to do with the type of angle difference (parallax) you need to create a stereoscopic effect.

0 upvotes
HowaboutRAW
By HowaboutRAW (Feb 29, 2012)

Odd that the review doesn't mention banding in large areas of shadow, nor is there mention of artifacts in photos with large area of skin tone. (I've seen both problems in a real world demonstration of this camera.)

Have these problems disappeared? Will DPReview be doing a fuller review?

0 upvotes
SamTruax
By SamTruax (Mar 2, 2012)

It did mention banding at the end. Noticeable at high ISO.

0 upvotes
J Parker
By J Parker (Feb 29, 2012)

Thanks for the review of an amazing concept. From a creativity perspective, the Polaroid analogy is definitely appropriate (just look at the revived popularity of low tech photography methods such as pinhole cameras and lomography). I'm looking forward to seeing what the naysayers of this camera say five years from now when the technology matures a bit. By the way, Facebook sized resolution is not exactly a limitation anymore -- the 2.5 billion photos uploaded to Facebook per month (that's about 964 per second) is just one sign of how the need to produce 'prints' of traditional resolutions might not be as crucial for many camera owners as it might have been even a few years ago.

0 upvotes
brendannicholas
By brendannicholas (Feb 29, 2012)

My photos can do that too! It's called photoshopping a photo taken at f8.

1 upvote
random78
By random78 (Feb 29, 2012)

No the result of light field process is very different from what are describing

3 upvotes
Joesiv
By Joesiv (Feb 29, 2012)

I've yet to see a good result with the photoshop technique of adding subject isolation. Care to share some?

1 upvote
Jogger
By Jogger (Feb 29, 2012)

i had fun playing with the embedded sample for about 30 seconds

6 upvotes
Faintandfuzzy
By Faintandfuzzy (Feb 29, 2012)

Considering all the cameras that could be reviewed...I'd rather see a review done on a pinhole camera.

3 upvotes
HowaboutRAW
By HowaboutRAW (Feb 29, 2012)

Is there a particular pinhole camera by a particular pinhole camera manufacturer you have in mind?

Here's an idea, you can make a digital pinhole camera by using a bare sensor from say an old mirrorless camera body, then you can blog about the results and post some of the results on Flickr.

NB: Search the terms "digital pinhole camera" on Flickr and check out the results. Looks like some are already doing such things.

Go to.

2 upvotes
HBowman
By HBowman (Feb 29, 2012)

Some ppl like to waste money in useless engineering. What a wast of time im dazeled.

5 upvotes
Goreyo
By Goreyo (Feb 29, 2012)

I wonder how many things in your home have their origins in 'useless engineering'.

13 upvotes
random78
By random78 (Feb 29, 2012)

Not really. The first commercially available DLSR had 1.3MP and cost $30000. But later versions built on it and look where the industry is now. This is a completely new technology and like all such technologies someone has to be the pioneer and get the first version out. The first iteration won't be the most exciting one but once it is in place it opens the door for more development to get it to a point where it will be more capable and more suited to a broader range of uses.

3 upvotes
HowaboutRAW
By HowaboutRAW (Feb 29, 2012)

The transistor would never have been developed if HBowman had a say, nor the hard drive.

It's amazing that when this guy Ng started out to learn this technology, no he's not the inventor, taking a picture required a huge wall of cameras and a super computer.

2 upvotes
reginalddwight
By reginalddwight (Feb 29, 2012)

Interesting concept. I'll pass and stick with my iPhone.

4 upvotes
HowaboutRAW
By HowaboutRAW (Feb 29, 2012)

The camera in the iPhone has an F/2.0 lens and you can refocus anywhere in the picture after shooting?

0 upvotes
Backpage
By Backpage (Feb 29, 2012)

Steve Jobs spoke with these guys about incorporating the technology into the iPhone....so, you never know.

1 upvote
HowaboutRAW
By HowaboutRAW (Mar 1, 2012)

Backpage:

The camera in the current iPhone people carry...

An AMOLED screen for a near future iPhone is more likely than a light field camera.

0 upvotes
eppicphoto
By eppicphoto (Mar 1, 2012)

The best part of the video was at the end when they did some decent high key lighting on the commentator.

0 upvotes
Total comments: 310
12