Henrik Herranen: Video of today is 4K / UHD, or in other words around eight megapixels. Lytro images show obvious digital artifacting even at 0.2 (zero point two) megapixels. Reference: pixelated out of focus areas between hand and background in http://www.dpreview.com/files/p/articles/1418225486/IMG_0331-ShallowDOF-520px.jpeg from article http://www.dpreview.com/articles/1418225486/lytro-introduces-focus-spread-feature
This technology doesn't work.
I've done a bit of 4K editing--just to try.
It's not impossible, but it takes still serious resources.
Everlast66: I am extremely grateful to its majesty, The Free Market, for sending this product where it belongs - history.It was obvious from the very beginning that it is not going to work - high price, inadequate image quality and questionable benefits.
But the most annoying thing for me was that the manufacturers were not open and honest for what they are selling. They are reluctant to tell their potential customers what the product is actually capturing and how. Many suspect this is because their product is just algorithms applied to regular image data. This would explain why they are unwilling to disclose any technical detail, because if they do they would not have a product to sell.
Proving you're not familiar with math.
Lytro has the entire market for plenoptic cameras, therefore Lytro has an infinite market share of that market. This is basic division. Here infinite market share means of course 100 percent of the entire market for plenoptic cameras. Everywhere you look for a plenoptic camera to purchase you'll only find Lytro.
You really need to think about things a bit more.
Indeed Lytro may fail, but the tech will live one in some other camera gear.
I'm quite certain that Mr Ng's PhD makes no speculations about a possible market share for plenoptic cameras.
Ng's PhD is in computer science, he wasn't working on some speculative fool B school proposal. In the late 1990s almost no one in B schools apprehended that search engines would be a really big deal for years to come.
Your comments read like much ill thought out B school prognostications. Such an irony that when Microsoft was being run by a business school grad he nearly destroyed the company.
neatnclean: I consider lightfield imaging as the most promising approach to eventually get cameras that no longer need large, heavy, expensive lenses. It may well turn out to be the most important step after replacing film with electronics to liberate stills and video imaging from the shackles of its 19th century beginnings.
Lytro is stil a long ways from that goal today, but they have proven it is possible. I salute them for this and i do consider those next 50M dollars to be a worthy investment to lay more foundation for my image capturing devices a few years down the road.
And lenses that change shape to aid in light capture.
Want to bet that there'll be a "discovery" that rods and cones in mammalian eyes have photo detectors in places were were told there are none.
AlexisH: I wonder what they'll try after video? Sounds like no one wants the current tech for stills. And then I expect that video will not provide the resolution for professional use and that the amateurs won't have the time and interest to select focus themselves.
Have you tried to edit 4K video?
I certainly don’t think the Lytro Illum is flying off shelves, but it’s only been out since the mid-fall of 2014. And it really hasn't dropped in price.
Some math for you: Lytro has infinite plenoptic camera market share, even if they’ve only ever sold 3 cameras.
If we, as in humanity, always chased “profit” within a few years, there’d be no internet, even transistors wouldn't exist. They were invented long before the official "discovery".
Forum comments here by those who’ve not used the camera and the software, that means the Illum, and at least skimmed the Ng PhD aren’t real serious evidence of lack of interest.
A lot what you’re written reads like attacks on Apple Macs in about 1996.
I’ve looked at Ng’s PhD, after I read it in some detail. And I paid to have it printed and put in a spiral binder–though I didn’t pay for every colour image–so that confused things a bit. In someways I’ve more interested in looking into the works cited in the bibliography.
bogdescu: Might it have dawned on them that merely postponing a decision (where to focus and what f-stop to choose) is not actually something desirable?
Well the surveillance Lytro camera, could be set to a lower gain (ISO) and the aperture could be left wide open, giving a faster shutter speed in addition to the aforementioned ISO thing.
Michael Uschold: Is anyone aware of a single person that is making serious use of this technology that goes beyond it being an amusing gimick? What would that use be? Is this person a lone wolf, or are there significant numbers of people that found a greate niche use for this camera? What would those niche's be? There was a lot of hoopla about the Segway, which many regard as a flop. It has had dramatically less impact than the inventors imagined, but I think there are some important use cases to justify continued production, e.g. large manufacturing floors, postal delivery and police officers.
Personally, I think Lytro will be very lucky to find any significant cadre of users for the camera, but people have boundless creativity, so who knows.
Ah, the real question. And a good one.
I've seen people using Segways, one last week. I've also seen electric unicycles, low-riding, which clearly use some of the same principles as the Segway.
Wing2: 50M is too little money for so much technical challenge the company is facing.
Probably need to output and store an enormous amount of data losslessly at video frame rate plus need to develop special software to display and edit the video with features and quality acceptable by professionals
Agreed, this is hundreds of millions of dollars project that will take 5 years to realize, that's with a good plan and the money to spend + advances in hardware.
HowaboutRAW: In 1974, maybe '75, IBM released it's first PC at a cost of something like $50,000, in mid70s dollars. (Irony the PC was at least assembled in Seattle.)
1981, different story.
Look what happened within 20 years to the PC market, and the usefulness of PCs. Every computer development for broad public use that has occured between 1995 and 2015 was well predicted in 1995 by anyone paying attention to computers.
Contrast 1975: Almost none of what came to pass in computers in the next 20 years upto 1995 was being predicted--even by people paying attention to computers. Maybe digital audio. Even into the early 1980s physicists who wanted to demonstrate a new program, software, they'd worked on at a conference would often travel to the conference with the program recorded on punch cards. That's cardboard.
No the Lytro is not likely to be as radical a change to photography as the broad use of a computer, but the nay-sayers here sure sound like many talking about computers in the 1970s.
Disagreement with is not the same as belittlement.
Nor is calling out preposterously made up claims about Lytro belittling the party doing the prevaricating.
You completely missed the point about PCs in 1974, and you're thinking of PCs over the last 20 years.
Your huge mistake: In 1974 everyone was saying exactly what you've just said of Lytro, but about PCs: "they're not a broad category".
Ah yes, the spurious more pixels makes for better photos line.
Retzius: Lytro is struggling because its a solution in search of a problem.
For the most part ice skating is not transport, though there are skates that double as cross country ski-like things that one can use on crusty snow and then skate along rivers with--but those are as obscure as the Lytro camera.
You've confused cross country skiing and ice skating.
No, the fact that people enjoy ice skating, and purchase skates does NOT disprove my point at all.
If ice skates had only been introduced to the world 3 years ago and no one were purchasing them, would you assume no one will ever ice skate? By your logic the answer is yes. Of course fact base reality has proven otherwise.
How is Lytro's Illum "indefensible"? Is there morality, yours, involved?
I take issue with the fact that you made stuff up to accuse Lytro of. And you've not retracted those claims, nor evidently made any effort to find the PhD.
Mr Ng is NOT on record as saying Lytro is not practical.
filmrescue: What I've always thought about Lytro cameras for photography...."Well that's really cool but I kind of know what I want in focus when I take the picture - most people do". Hope they have better luck with video...it actually makes a lot more sense. Focus pulling in post would be really useful.
And hundreds of millions in losses too.
Scales USA: All they need to do is purchase it!
It's expensive for a collection of early photos--not recording some remarkable event.
Some UK library may very well be looking for the monies to purchase the photos during this delay.
Big libraries, and museums, even very wealthy ones backed by government funding, all have limits to their budgets.
Perhaps the British Museum is looking for the money in say next year's budget.
It's spectacularly ironical that you've still not read the PhD, and seen where the data is coming from.
Hint: There's a huge hardware modification to the sensor, you've totally ignored that. There's also significant maths explaining why the data should be there to read, if in fact any of the equations used to model light are close to accurate models.
And the fact remains that you still wrote above this piece of ignorance:
"Many suspect this is because their product is just algorithms applied to regular image data."
None who are honest suspects any such thing.
I'm sure there are cases where I've misread something in a comment and commented wrongly myself--and I try to note those mistakes.
Unfortunately, for you, that aint the case here--I've read what you've misrepresented.
Because you took an optics class doesn't mean that you understand how this camera works.
Why do you keep wanting to dig yourself in deeper--you're not going to fix this until your read.
In 1974, maybe '75, IBM released it's first PC at a cost of something like $50,000, in mid70s dollars. (Irony the PC was at least assembled in Seattle.)
grasscatcher: I have to wonder, with the improved processing speed of modern cameras (e.g. - 30fps @ 8mpxl), could systems be developed for focus bracketing, much like what is done for exposure bracketing? Thirty pics taken, each with focus pulled in a bit from the previous exposure, then a final processed pic with relatively infinite focus (algorithm could be developed to focus from, say, 30' to 3' based on x focal length, 50' to 5' for a longer focal length, etc. to more accurately capture moving subjects, plus a setting of 1' to infinity for landscape still photos...).
This would be a great boon for those of us who require maximum DoF for certain photographic applications.
Pretty sure focus bracketing already exists on some conventional cameras.
It will take some extraordinary software to sort out what is in focus and then discard the rest--easier with a Lytro modified sensor. Then Lytro can still create a shallow DoF image where it would be next to impossible with the image generated with bracketed focus.
Then the elephant in the room, similar to the moving sensor of the latest Olympus EM 5II:
The Lytro can fire the shutter once at say 1/250 of a second and capture that image data. While a bracket generated image couldn't really be used where there's any significant movement in the frame.
Lytro, like ice skates are the tech, both allow things not possible with previous tech--winter boots or conventional digital cameras.
Ice skating still offers no solutions to a previous problem--Lytro is different in this respect. And then like skates plenoptic photography opens up new possibilities.
panpen: 4MP camera? What is this? 2001?
Does a few things my Canon G2 doesn't.
And the 54MB files are a bit bitter than my G2's raws--which are 2MB.
Eigenmeat: The sad thing is that a MISS FOCUSED shot from a decent P&S/phone still has MORE resolution than an IN FOCUS shot from a Lytro.
And the sadder thing is the mega pixel race showing up in these comments.