Camera Crew

When Ren looks up to the sky out of the corner of his eye, you know he has something cooking.  Four years ago, he was working on the menu for a gourmet meal… Kinda hokey, but we thought it might be fun to share our story from the very beginning – a little primer on how to start a picture revolution.

Back when our founder Ren Ng was still a PhD candidate working on his dissertation at Stanford University, he saw the enormous potential for light field cameras. He also realized if he wanted to succeed at making a business out of his ideas, he would need a team with unique skills – plus, a triple-shot of persistence.

Ren’s Original Believers

Just Chillin' - Ren's Original Believers - Tim, Colvin and Alex

Colvin Pitts, Tim Knight, and Alex Fishman made up the initial team of engineers Ren hired at what was then called Refocus Imaging.

Colvin came on board first.  He had been Ren’s undergraduate roommate, but it was over dinner a few years later when Ren showed him the technology he had been working on.  “Needless to say, I was blown away by what could be done.  At the time, I was in a good position in another startup, and I had no plans to look around or leave.  However, the opportunity to help develop this truly disruptive technology was more than I could pass up.”

Next up, Tim. Fresh out of Stanford, Tim has a PhD in computer science. “I was new to light fields, computer graphics, and the inner workings of digital cameras, so there was quite a learning curve for me at first. It was incredibly fun and satisfying.”

Rounding out that first crew was Alex, who came to the team with deep experience in the digital camera industry. “Honestly, I was skeptical. Then Ren showed me the actual camera prototype in action.  And here I was, standing in front of a young, well-spoken entrepreneur, who breaks the laws of physics as we knew it.  Right there, right in front of my eyes.  I was stunned. I’ve seen cameras evolve from the expensive, bulky and low quality devices in mid-nineties to a very mature and saturated market, where innovation was nowhere to be found. The camera companies tried to differentiate themselves solely in shape and color of the camera and how many faces it would detect. Back then, like now, everyone was desperately looking for innovation.”

How did you decide to build your own camera?

Alex sums it up this way: “Breakthrough after breakthrough, the technology reached a point where we could show an amazing and engaging experience to capture and share an image. The speed our prototype would take the next picture was indescribable. I could not believe my eyes! And then sharing the pictures with my friends was another big wow. They all loved the pictures.

But the single recurring question was, ‘When can I take a picture like that?’ This was about the same time when we reached the conclusion that if you want something done right, you ought to do it yourself. And so we did – we decided to make our own Lytro branded cameras that will forever change the way everyone takes and experiences pictures.”

What were some of the hard parts?

One challenge for the initial team was to build prototypes to demonstrate to outsiders, potential investors or partners, that it really was possible to miniaturize the technology to make a consumer-friendly camera. “I broke a lot of prototypes in the early days, but some of the ideas we developed at the time have been greatly leveraged in the transition out of the lab to mass production,” shared Colvin.

Tim remembers, “We worked very closely on improving the core light field technology and theory, on image processing, and on software on many platforms, including all the demos that were used to show what we could do – web, PC, mobile, 3D, on-camera software, etc.”

What keeps these guys excited about light field cameras and building a new business from the ground up?

Tim says, “Working on light fields has been a great experience. It is a challenging and fun technical space with a lot of meaty image science within it. It’s exciting to bring this technology to the mainstream consumer.” For Alex, it’s the journey ahead. “Light field is a very powerful concept and many new discoveries are still ahead for us.” For Colvin, “What excites me the most is this is the first camera that creates an experience suitable to the digital age we live in.  Photography has produced static images since its inception, and finally we are creating light field cameras that can create living, interactive pictures.  It is really exciting.”

 

 

  • Average Joe

    You really got to improve your R&D funding. In this era, there’s no practical purposes for a camera that delivers barely 1 megapixel photos. We’re in 2012, not 1998(where it’ll then fit with Kodak DCS). My advise to Lytro is to develop the next model with more conventional design shape and at least 5 megapixel sensor even before this one hits the market. Otherwise i’ll probably be seeing this camera used in passport photo booths or by kids who love novelty devices.

  • egorbogat

    Hello to all!
    Could you be so kind and help me?
    I am working with Lytro camera now. I took my sample’s picture. After I click by different object on my pict and save jpg-format pictures. I click by objects with a step 2 mm (in deep, my sample – is periodic black lines in white background). I have received same pictures even if I clicked in different objects. Is it means that Lytro camera (refocusing algorithm) has special refocus point??? How to identify these refocus points in camera? I didn’t find answer in Ren Ng thesis. Please, could you help me!
    Thanks a lot!

  • Pingback: Our First Photowalk with Lytro Camera Owners | LYTRO BLOG

  • Jim

    Since all the info is in the picture why not have it all in focus at one time? Printable?

    • Lytro

      More info on both of these, which are possible, in the FAQs on our Support site. Search for “printing” and “all in focus” or keywords for any other questions you may have.

  • Roy Hackett

    Seeing all of the patent battles between Apple and Samsung over the general look and layout of the iPad as first seen in the movie 2001, I am concerned for you folks based on the movie Blade Runner. As I recall, Harrison Ford loaded a still photo into his desktop and instructed it to move a box around the image until he found a mirror reflection. He then asked to zoom in and refocus. The tech made an impression on me and still does. One on order for me. Good job.
    Roy

  • Pingback: Lytro Light Field Camera lets you focus after taking your shot « Kuwait Info.

  • Pingback: Canon still mirrored

  • http://hirgereterfer.org Giuseppe Duell

    It is great to see scenery which make perception and are beneficial anyway.

  • http://forum.inphase.ru/index.php?s=c26c80316aab2da935de08cef55da1cc&showuser=111772 Ahmed Rumer

    Recommended this to my Facebook friends, they seem to love your website! :)

  • Andres-san

    John Helms Posted a link to here and already im Pumped for this New Camera to Come out in the U.S!!!! Also Im glad to hear the prices wont be Jacked all the way up!! :D

  • http://www.facebook.com/media/set/?set=a.323455428344.194547.323330173344 Mark Kane

    I haven’t read the dissertation but the main innovation must be in the chip, specifically the ability to the angle of incoming rays. This interests me a lot. My current dream project is suspending a gyro-steadied DSLR from a sort of cantilevered mini-roller coaster and recording the journey of a digital bee from first sight of a distant poppy flower into the private world of its center. A lot of the mechanical challenges would go away with a light-ray DSLR. And new photography would be feasible.

    • http://www.facebook.com/media/set/?set=a.323455428344.194547.323330173344 Mark Kane

      …”to record the angle of incoming rays.”

  • http://yugiohgx.net Yu-Gi-Oh GX

    *I used to be very pleased to find this internet-site.I wished to thanks on your time for this glorious learn!! I definitely having fun with each little little bit of it and I’ve you bookmarked to check out new stuff you weblog post.

  • Paul Cuddihy

    I wonder if anyone has looked into the range of possibilities this technology could have in terms of remote sensing for GIS data.
    It could make many of the existing, and expensive, data gathering methods redundant.

    The potential for this technology is mammoth if its developed carefully.

  • SMcQ

    Your camera hasn’t even come out yet and already I’m thinking of a hack. Terrible presumption, what can I say?

    This what I have in mind, inspired by a comment Ren made in the CNET interview. At the UC Davis Tahoe Environmental Research Center, Incline Village, NV, we have in our exhibit area a stereo microscope set up to view live plankton. There is a monitor which taps one view. There is no way to view stereo except through the two objectives, which kids have a very hard time doing.

    I’m thinking that if I can configure the Lytro camera with a powerful enough additive lens (doubtlessly the camera lens is not interchangeable), and the software allows the selection of stereo pairs from the light field (which clearly it does), then we could feed a standard signal through HDMI 1.4 connection to a passive 3DTV monitor (LG or Visio) for the whole class to see in stereo 3D and deep focus.

    So what I’m asking is, would this violate the warranty? (that was a joke). No, what I’m really asking is, can I tap out from the camera or the software on a PC a stereo pair that can be read by a 3DTV? It would be most excellent if the camera could be tapped without intervening hardware.

    SMcQ

  • SMcQ

    I want to submit to Lytro an array of CGI rendered synthetic microlens images that you could synthesize using your software.

    What is the optimal pixel dimension for a microlens image? What is the optimal array configuration for a microlens set?

    My approach, subject to advice, is to render a video file of the microlens set. This is easiest to render as it simply requires an animated camera motion of a static scene, so that the set represents a raster of microlens images. But this might not be optimal for processing.

    An offlist response is fine. I realize there may be some disclosure issues in answering this message.

    SMcQ

  • SMcQ

    Martin,

    Not an expert, just a lowly science writer/illustrator.

    You might be right, but from what I can discern, even the compound eyes with focusing elements are used to compose mosaics from the very narrow angle of view afforded by each element. The elements of this type are in a sense modified light pipes with some imaging capability per element.

    The best candidate for what you are surmising would be the “Apposition-Neural Superposition compound”. Read about it here, section E:
    http://www.eyedesignbook.com/ch2/eyech2-def.html

    As far as I can tell, recording a lightfield using the Lytro method requires each element, or microlens, have a wide field of view and sufficient pixels for coding the direction of incident light. I’m not sure a primary lens is as much a requirement as a convenience.

    I don’t think any insect eye, with or without a “lens” has elements with wide fields of view and many “pixel” receptors for each to record the multiple views, enabling the insect neural system to process a lightfield. I could be wrong.

    I leave it to you to research further and become the true expert instead of an intellectual gadfly like me.

    SMcQ

    • CC

      From the paper it appears that this depends on the lightfield representing the path through two planes – one of which is the aperture of the primary lens – thus I don’t see how this would function sans such lens.

  • Joan Robinson

    The technicalities are way beyond me (I’m an ancient old thing and truth be told, not in the right intellectual stratosphere) but I am fascinated by what you have achieved and can’t wait for an opportunity to see the finished product and provided I can afford it, buy one. (I’m not a really good photographer but I really enjoy trying and who knows, your invention may help to make my efforts look better………….) I hope it will be available over here in the U.K.

  • Cindy Maure

    An absolutely fascinating technology. I am so pleased to hear that you intend to keep the cost of such a camera at a modest consumer price so that many will be able to take advantage of your new product. When I think of the technology, I am reminded of holograms. I wish you success with this endeavor.

  • http://www.adabadadingo.com Harold Sitler

    Hello There. I found your blog using msn. This is a really well written article. I will make sure to bookmark it and come back to read more of your useful info. Thanks for the post. I’ll certainly comeback.

  • Joe

    “What excites me the most is this is the first camera that creates an experience suitable to the digital age we live in.”

    What excites ME the most is this is the first digital camera that truly rivals a film camera.

    Great work!

    • CC

      This is a really, really fundamental change, in that photography has cascaded from the invention of the Camera Obscura whose basic principle was to recreate an image at a focal plane through the use of a lens to bend the incoming light.

      Lightfield photography takes a new tack, to record the field rather than a fixed surface, something that was not possible with the mechanisms possible pre-digital-age IMO (except for holography.)

      The consequences of shifting to a lightfield capture maybe even more dramatic than are currently realized, Still, even the list articulated in Dr. Ng’s dissertation and in other research are high value.

  • http://www.Outpost81.com Robert Trudell

    Light Fields are great, now I’m wondering if Sound Field recordings can be added?

    –Rob

  • Paul

    Have you considered an entirely new “focus” method? For example, currently you (and all other cameras) are focusing on a plane perpendicular to the axis of the lens of the camera at a distance of X feet or inches… you do it dynamically, other cameras do it as fixed. But if your technology supports this, what about rotating that plane in any combination of the X, Y, and Z axis at a distance of W feet or inches (performed in software). That could be rather interesting.

    In addition, and as a requirement for the above, after viewing and playing with multiple of your sample files, how many ‘focal distances’ (for lack of a better term) are there? It doesn’t seem ‘infinite’, as a matter of fact it seems like there might be three or four. Which is fine, I am just curious.

    Last, what will be the minimal focal distance of the camera? Will it have removable lenses?

    Good job guys!

    • Fritz

      There are cameras who can rotate the focal plane. We call them a view camera or technical camera. They have been around for some 80 years. It works by tilting the front lens of the camera. Canon sells lenses that can tilt for their reflex cameras. You can probably tilt the focal plane with a lytro camera using software. Or even bend the focal plane: I’m sure we are going to see a lot of exciting developments with this technology.

  • http://www.rickkinnaird.com Rick Kinnaird

    My dad was one of the first dozen people hired by Dick Perkin of Perkin Elmer (Hubble Space Telescope, U2 Camera, Schmitt-Baker Camera, etc.) Iwish he were alive today. He would have loved to hear about this technology. Before hiring on with Mr. Perkin he had talked with Dr. Land who was fooling around with some polarizing disks. Dad laughingly used to say that he saw that wasn’t going anywhere.
    Good luck guys and I can’t wait to get a camera, if for no other reason than to connect with my dad.
    Thanks,
    Rick KInnaird

  • John Christiansen

    Been in photography,schools and all!
    Started in1959-I am 71 yrs. old!
    I am a Canon enthusiast from the model 2000 and up!
    “I AM EXCITED!” which is rare for me!
    Can’t wait!

  • G.Zink

    Exciting/outstanding creations. If you keep the camera at about $400 like the news stated, what will be the cost of the software to
    create the ‘magic’ images?

    • Daniel Tuck

      I would have hoped the software would be included, otherwise it would be a bit useless without

      • Lytro

        Our first light field camera includes a light field engine (processing software) that lives on the camera itself, in desktop software and travels with the living pictures as they are shared online.

  • http://marco.guardigli.it Marco Guardigli

    you are great!

    My compliments to the inventor, for the great ideas, and best wishes for a successful business future to the whole team!

    Next steps will be binocular cameras, 3d pictures, varifocus movies, and of course image synthesis, so that you can take a picture from one point and then ‘move’ also the viewing point after picture has been taken.

    Keep me informed about your ideas. I also have many interesting potential applications for automotive industries, such as scene analysis for computer vision, obstacle detection, and in general assisted driving.

    do your best to improve the world !

    Marco

  • Nicholas

    I am wondering how you can see the picture on the back of the camera if you focus it after on the computer. Or, do you select the focus point you want on the small uncalibrated camera screen just to see what the picture looked like?

    • Lytro

      You’ll be able view your picture and adjust focus on camera, on your computer and online.

  • Don Richardson

    It would be my dream to share in the seed capital of Lytro. After Stanford (BA Econ 59 and MBA 61) I was in Peace Corps, Citibank, years of Govt. service with USAID in many countries including Iraq and Afghanistan.

    Can I invest in Lytro before an IPO?

    • Marge

      Atta boys!

      • Drew

        Me 2, how?

        • Jeff Gammill

          I am also interested, should there be an IPO opportunity.

    • Ahmed Elalfy

      I’m an engineer / investor, very interested in this..

    • Albert

      Ditto

  • Martin Ross

    Do you think insects, with compound eyes, might be doing something similar and actually have better vision than biologists think? Maybe the algorithms you use to construct a picture from hundreds of partial images are mirrored in the brains of insects. They have thousands of receptors, which, if treated as single optics, limit them to a few millimetres of distance vision; if acting like a light field camera maybe they see quite well. They fly so quickly it seems 3mm would not be enough.

    • SMcQ

      Insect vision has been well studied. The compound eye of an insect is more like the retina of an advanced eye, flexed to convexity so that it gathers light from the environment rather from an optical lens. Each lens-less element has a field of view limited to its own particular sector of the entire potential view, almost all around the bug. Each element performs as a light pipe, conducting photon energy incident upon it from its narrow field of view to sensor cells. The compound effect is a mosaic, much like a tile mosaic, of the light intensity at each element. Insect vision is optimized for quickly detecting movement, not for sharpness of details.

      By contrast, each micro lens in the Lytro array, as I understand the technology by reading the dissertation and examining the illustrations, is a true optical lens that captures a wide field image focused upon the number of pixels available to each microlens. Because the light from each microlens is distributed across the known geometry of the sensors, the input directions for each micro lens can be sorted and assembled in the aggregate by computation.

      So this is not a regression to compound insect eyes evolved to detect motion almost 360 degrees around, but little else; it is a synthesis of thousands of optically advanced eyes that can record the field of light passed through a prime lens, and make this data available for selection through computation.

      In short, insect eyes don’t do light fields, they do movement fields.

      • Martin Ross

        Are you an expert? From what I have read some compound eyes have lenses which form multiple images.

  • Don

    Would it revolutionize not only landscape, portrait, macro photography but also wildlife photography where people need powerful lens like 500, 600 mm ?

  • Ólafur

    I see your demo pictures and am impressed. I would like to know if you can select the depth-of-field.

    • http://johnhmoore.com John H. Moore

      Same question: with adjustable focal point, what determines the depth of field? Is depth of field also adjustable after the fact, or is it determined when the shot is taken? What happens to the concept of shallow depth of field bokeh… can you get a beautiful background blur with this technology? Thanks!

      • http://johnhmoore.com John H. Moore

        OK, I’m through Chapter 3 of the dissertation… is this camera a fixed f/4… optically? …but then you can use software to decide the areas that you want in focus in the final image? So the bokeh will be that of the main lens, at whatever f-stop is required so that the final f-value is 4? If you put a larger than normal for f/4 area into focus, then leave the remainder out of focus, does the gradation from in-focus to out-of-focus have a different appearance than we’re used to with traditional cameras? …wish I had one to play with!!

        • steve

          By definition, a light field capture device captures light rays as individual vectors. Focus and DOF is not a part of this, they are decided during the “flattening” process after the image is captured.

          Though, there will be limitations in the capture device itself (the density of the microlenses used on the Lytro will determine the range and granularity of the captured light field), so focus and DOF of the lens may come into play deciding how much refocusing/DOF adjustment you can do.

          • http://johnhmoore.com John H. Moore

            Is there a relationship between the number of directional resolution microlenses at each (x,y) cluster and the ability to later refocus? Is refocusing dependent on there being an appropriate (u,v) point at/near where you want to refocus? …such that the dissertation prototype would have had 144 potential focus points, but not an unlimited set?

          • http://davidgerhardart.com David Gerhard

            If information is saved as vectors, does that mean that printing size can be blown up to large scale? As an artist that is an additional benefit onto the growing list of your camera. Cannot wait to use one.