PetaPixel

Diving Into the Tech Behind the Lytro Illum and Its Impressive 30-250mm f/2.0 Lens

Lytro_2

Lytro came into the photography world not only to create a novelty product, but to fundamentally change how we approach image capture. Because despite light field photography being around for over a century, it’s only with the latest technology that the company is able to exploit what it is a camera is truly capable of doing.

We recently spoke with Lytro about its upcoming Illum camera a bit, diving into the technology behind the specs and revealing how Lytro’s approach is allowing the company to not only step, but leap into the future.

When Lytro started the research and development around its Illum camera, almost the entire design was based around one component: the glass. Despite the fact that the sensor and accompanying software are the main pieces of technology Lytro has up its sleeve, with the Illum, Lytro wanted to make sure that the lens was truly capable of letting the sensor shine in an effective form factor.

Lytro_4

The irony is that the sensor would dramatically affect how the light field camera company went about its lens design. Being a light field camera, the sensor captures data very differently than a conventional camera setup, which in turn allows Lytro to approach lens design and manufacturing unlike any other camera company out there.

Rather than capturing only the intensity and color of light that hits each pixel, Lytro cameras also capture the angle at which each photon arrives. Made possible with the hundreds and hundreds of microlenses in front of the individual pixels, this extra data gives Lytro more control over what it is a lens is capable of compared to conventional cameras.

AC5F885BD724914EB13A9C0F8553B52C@SDIHQ.com-LYTROILLUM_LensCompare

As we’ve noted before, the Lytro Illum features a 30–250mm equivalent f/2.0 lens. Yet, as you can see in the above comparison, the camera itself is less than the total length of a conventional 70–200mm f/2.8 lens. In fact, not only is the entire camera smaller than the 70–200mm f/2.8, it also comes in at a lower price-point — $1,500 for the Lytro Illum vs. ~$2,500 for the 70–200mm f/2.8 with image stabilization.

So how is it possible to get a faster aperture over a larger focal range, all at a lower price point? Well, this is where it gets a bit confusing. To boil it down into one word though, the answer is magic math.

Taking into account the capabilities/limitations of its current sensor and microlens technology, Lytro was able to go about constructing the 30-250mm f/2.0 using almost-off-the-shelf optical components. Its proprietary lens design would result in absolutely horrible photographs on a conventional sensor, but the capabilities of the light field hardware and software behind the Illum allows them to effectively compensate for this.

Lytro_3

Lytro is basically turning physics into software. By creating its own lens designs alongside its sensor technology, Lytro can appropriately map out where every ray of light is coming from when it hits the sensor. Then, rather than relying only on physics (as essentially every other camera does) Lytro can re-arrange those rays of light in order to end up with a corrected picture, and in so doing eliminate the physical restrictions that lenses such as the 70–200mm f/2.8 are built around.

Through this methodology of reconstructing an image from the data captured by the sensor, Lytro is able to bypass the optics that currently limit the size of lenses and cameras. Plus, as the price of optical components goes down and the computing power capable of being put into a camera increases per Moore’s Law, Lytro will continue to benefit.

This gives Lytro a major advantage in the photography market. Rather than owning any of the physical components itself, the only technology the company is directly responsible for improving is the software and algorithms. Thus, as the third party manufacturers of optics and computer chips improve those components, the off-the-shelf components Lytro uses will become higher quality, all the while staying affordable thanks to economies of scale.

Lytro_1

Light field photography has been around far longer than Lytro has. In fact, Lytro isn’t directly responsible for inventing anything revolutionary… the company is simply putting together a number of technologies that have already existed.

But by properly implementing and exploiting these technologies to squeeze out every possible photographic advantage, Lytro is destroying the boundaries of photography as we know it.


 
Get the hottest photo stories delivered to your inbox.
Get a daily digest of the latest headlines:
  • james

    Lytro suddenly became interesting!

  • http://500px.com/mlianopr mlianopr

    Take that, Sony a7s….. lol

  • http://transienteye.com Mark

    Rather than relying on Physics? So what does it rely on then – magic?

    The reality is that this is effectively a 3.2x crop factor camera with an equivalent still picture resolution of 4MP. While not constant aperture, Sony are busy making smaller f1.8 capable lenses with that work with a larger sensor (i.e. RX100).

  • OtterMatt

    Didn’t we JUST get done with a whole series about how the crop factor makes these lenses NOT equivalent to an f/2.0? I feel like we need another rating, like maybe how we rate computer monitors based on contrast ratios. I have a hard time believing that this thing will compete with a DSLR, no matter how they’re bending science to do it, especially with secondhand components.

  • jon

    This is still an f/2.0 lens. It just isn’t a 30-250mm lens. Why can’t people understand this?

  • Bryan

    f/2.0 will always be f/2.0 regardless of crop factor!!!!

  • Kynikos

    Another hyperbolic Petapixel adjective:

    this lens isn’t “impressive” at all.

    Next article: Gradma takes awe-inspiring photograph of her sleeping cat.

  • Kynikos

    It relies on… Megarays! From the planet Mongo!

  • mikeaubrey

    Who said it was equivalent to an f/2? Show us where.

  • Axelanti

    only the DOF changes, the luminosity stays the same

  • doUhaveANYtobacco

    Ahhhhh, so wonderfull, I see now… so they actually use little tiny pieces of glass inside the magic object to be able to seed the fields, quell bonheur!!! Maybe it’s time also for me to pretend I now a secret, hi hi hi

  • Ben

    What a great press release!

  • Rob Dickinson

    Smaller than a 70-200 f2.8, but its only an ‘effective’ 30-250mm. In reality its what 10mm-80mm, doesnt seem so compact now?

  • Rob Dickinson

    It relies on trying to convince the consume this is a 30-250 lens, its not. Writing effective in really tinny letters and hoping no one notices…

  • OtterMatt

    It says it in the headline, in the article and on the side of the freaking lens. Read.

  • OtterMatt

    Okay, yes, technically it is compared to it’s own focal distance, I get that much. But it’s still shady marketing, because it just won’t look like even an APS-C f/2.8, much less a full-frame one. And since one of the highest advantages of a DSLR is the improved contrast, color depth, and detail, I still feel like rating a lens like that based on f-stop is misleading.

  • jon

    This is an f/2.0 lens–that is a literal fact and cannot be misconstrued. The lengths are full-frame fov equivalents. The sensor is small. Why are you equating the iris with the depth of field? There are two other variables to that equation. Why are you even worrying about depth of field? That’s a small and immature way of looking at the art.

    And what is there besides technically correct in this situation? It’s not misleading marketing. The only thing that might be misleading is the lens zoom range. It should say “equivalent” right beneath it. Everyone does it because it’s easy to understand the usefulness of the camera/lens when speaking in focal length equivalents.

  • http://transienteye.com Mark

    It is misleading marketing because it is being explicitly compared with a full-frame 70-200 f2.8 IS lens, which gathers much more light due to the longer focal length (f-stop is a ratio…) and the need to illuminate a larger sensor.

    Lytro is destroying the bounds of marketing hype, not photography.

  • jon

    That’s stupid. Every bridge camera or superzoom uses the same equivalency numbers. Don’t be thick.

    And the lens still illuminates the area of the sensor with the same intensity. Don’t act as if it doesn’t. If you want to shoot with your small sensor camera in the dark at f/1.4, it will be just as bright as your full frame camera with a lens at f/1.4 (given the same iso and shutter speed and t-stop).

    Such a silly distinction. Completely unnecessary.

    Also, why are you commenting? I covered your comment in the comment you are responding to. What?

  • jon

    everyone does this. including sony with their rx10 and rx100. it’s a matter of convenience. obviously it’s not a 30-250mm lens, but it helps camera users understand its fov range.

  • Gvido Mūrnieks

    F/stop is a measurement unit, that is used for calculating depth of field, not exposure. When you are talking about light coming through the lens, to sensor/film – You must use T/stop as measurement unit. What do we get out from this? When we talk about FF lens equivalent lens of Lytro’s lens – it would be: 30-250mm f/4.6, t/2.0!

  • Oskarkar

    There is something strange: if one is recording the phase of light(angle is not sufficient to make focus) then one could make not only variable focus but also 3D images out of it. Phase light recording is a difficult task and involves at least one reference. I guess in reality there is no such “angle” recording but they record sort of video with focus ramp and their software lets you choose only one frame with the focus you want. I still appreciate the lap into something new no matter the way they do it.

  • Gvido Mūrnieks

    This is why crop factor applies not only to focal length, but to f/stop too. That’s why, when you put lens on different sensor sizes – both, FOW and DOF changes.
    When you say, that luminosity stays the same – that means, that T/stop doesn’t change. For example, whe you put 25mm f/1.8, t/2.2 lens on MFT, that has 2X crop – you will get 50mm f/3.6, t/2.2 equivalent.

  • lillianhoward

    Michael .
    I can see what your saying… Carolyn `s stori is incredible… on wednesday
    I bought themselves a Renault 4 since I been earnin $5466 recently and-a
    little over, $10,000 lass-month . it’s certainly the most rewarding I have
    ever had . I actually started six months/ago and pretty much straight away
    was bringing in more than $74.. per/hr . over here

    w­w­w.W­­O­R­K­S­­7­7.c­o­m

  • PatsyWerstleritu

    My Uncle
    Riley got an almost new red GMC Canyon just by some parttime working online
    with a laptop. visit their website F­i­s­c­a­l­p­o­s­t­.­C­O­M­

  • Patrick Griffin Santucci

    Sigma is the only one with a 1.8 lens, not sony.

  • jon

    Depth is a side effect of a lens’s f-stop, lens design, and focal length. F-stops are used to calculate exposure, first (unless you’re using something like Minolta’s 135stf). How do you suppose you might use a light meter with this Lytro camera if you adjusted the F-stop to f/4.6 when it’s really an f/2? The light meter wouldn’t make sense, then, would it? And of course the transmission values are slightly different, I already stated this. But very few photographic lenses use t-stops. That’s why you calibrate your light meter to your camera and lens.

    And no, this is still an f/2.0 lens. Not a t/2.0 lens! It’s very likely a t/2.4 lens or so. But it is 100% an f/2.0 lens. It’s just not a 30-250mm. If you adjusted the f-stop for depth of field, light meters would be useless.

  • Jason Mayfield-Lewis

    Yes, f/2.0 is f/2.0. Just like a DX 35mm lens is still a 35mm lens, regardless of the fact that you’d need a 50mm lens on FF to have roughly the same AoV. But let’s assume that you have two sensors of the same underlying technology. One is FF, and one is APS-C. To the APS-C sensor, you attach a 35mm f/1.8. And to the FF sensor, you attach a 50mm f/2.8. The AoV, DoF and low-light performance would be identical (or near enough, excepting the difference in the proportion of each pixel’s surface area that can gather light as the pixel pitch becomes smaller and resolution increases on a given size sensor.) The surface of the either sensor has to gather the same total light to provide the same image brightness at a given gain. So per area, the APS-C sensor would have to gather double the light since it has half the surface area. Obviously, the cameras’ ISO equivalents are calibrated in such a way that a given ISO speed will result in the same brightness of exposure for the same exposure length at the same aperture, regardless of sensor size and capability – but all this means is that a given ISO-equivalent is ~2x the electronic gain on the APS-C sensor in my example (which should approximately half the signal-to-noise ratio – i.e. double the noise.) Since most consumers wouldn’t know how to begin to understand this, is it not somewhat disingenuous to market camera equipment on the basis of these numbers? IMHO, we need a ‘clean sheet’ system of providing equivalent focal length, image brightness and DoF relative to sensor size… It would help people understand the capabilities of what they’re buying and stop all those ridiculous hypothetical flame-wars about which system is better or worse numerically (of course I would argue that the best system, no matter what, is the one you are happy using – but that’s a different conversation altogether.)

  • Gvido Mūrnieks

    You are right and theoretically, I agree with your your points. And I understand, that f/value as measurement for light field camera is useless.

    My point is, it would be a lot more convenient, to separate t/value, from f/value, when talking cropped cameras, teleconverters and speed boosters.
    For example: Nikkor 50mm f/1.4, t/1.6 on DX body will be 80mm f/2.2, t/1.6 equivalent.
    But, when we add speed booster, between lens and DX body, it will become 50mm f/1.4, t/1.1 equivalent.

  • jon

    I am not only right theoretically, I am right technically and practically.

    If you really want to normalize all the oddly-sized sensors out there, then we should discuss lenses in terms of field of view instead of focal length. Otherwise, don’t be so tacky as to insinuate that the most important quality of a lens and camera system is the thinness of its depth of field.

  • jon

    This is wrong, then. Depth of field remains exactly the same across bodies, even if the field of view changes. F-stop remains exactly the same across bodies. T-stop remains exactly the same.

    The lens never changes. There is no such thing as a crop factor. A 50 looks like a 50 regardless of sensor. The compression of space is the same, regardless, and the qualities of the lens still translate across different sensor sizes.

  • Rob Dickinson

    Yes everyone does that, but no one crows about how small their lens is in comparison to a full frame lens. Thats the bit that is very bad form, they are using a tiny sensor.

  • jon

    Still can’t be adjusting f-stop based on crop factor. I’d prefer manufacturers denote the actual focal length, too. But then they’d have to denote fov and consumers would get confused. Is anyone really expecting their tiny rx100 to look like 135.

  • Aiden

    Nope, you’re wrong. Sony has at least a few f/1.8 E-mount lenses available already.

  • jon

    the f-stop stays the same, too. f-stop doesn’t measure depth of field in the way you describe. it measures the diameter of the iris to the focal length of the lens. it is a resolute measurement.

    you’re confusing field of view with focal length. out of focus areas have more to do with sensor to subject distance and focal length than with anything else. aperture can control that. field of view should not be a part of a technical discussion about lenses.

  • Rob Dickinson

    I didnt mention f stop. This is still an f2.0 lens regardless of the sensor size, but a small sensor does make it a more compact lens ( no need for as large an image circle) , and has many drawbacks for noise and dof sure.

    But you cant tell people how clever you are that you have made a 30-250 f2.0 lens smaller than a 70-200f2.8 lens when in fact you have actually made a 10-80mm lens.

  • jon

    Oh, then I am in complete agreement with you.

  • stubaw

    This part of the article stuck out for me as well.
    I can only guess that the author intended to say that part of the process (focusing) is done in software.
    Doesn’t sound right though.

  • Chito

    You know what, I’ve gone through the whole spectrum of emotions regarding this.

    We’re just arguing semantics but the basic truths are:
    a) They want you to think that the lens is just like a big expensive telephoto lens.
    b) The image won’t look like a photo taken on a big expensive telephoto lens on a full frame camera.

    I’ve basically settled on, if everyone else is doing it, why wouldn’t Lytro, as a company, label their lens as a 30-250mm f/2.0 lens? I mean, I think anyone who actually knows what a crop sensor and f/2.0 actually means knows what is going on and what this implies regarding depth of field. And anybody who comes up to you with a superzoom and thinks you’ve wasted your money on a white L lens obviously doesn’t know what he/she is talking about.