Lytro "Light Field" Camera: Not the Next Revolution in Photography

Some pundits claim that the Lytro "Light Field" camera is going to revolutionize photography: Here’s why they’re wrong

Has it hit the market yet? The fact that I’ve heard neither cheers of delight nor groans of disappointment seems to indicate its presence in the marketplace is pretty what I expected in terms of impact, which is to say… minimal.

For those who don’t know, the Lytro uses what’s called Light Field photography to allow you to choose your plane of focus after you’ve taken your photograph. There’s no focus control on the camera; you just point, click and decide later if you want the foreground in focus and the background blurred or vice-versa. Interestingly, you can’t have both in focus – the company says they’re working on this but there’s no indication of when it will happen.

The problems? There several. Let’s start with the most obvious one: Tiny image size.

The images produced by the Lytro are approximately 1.2 megapixels in size. Since the 16GB version of the camera stores 750 photos simple mathematics tells us the images are around 21MB each. 21 megabytes for a 1.2 megapixel photo is a pretty steep storage price to pay. Early speculation held that Lytro technology required a 64 megapixel sensor to create a 1 megapixel "light field" image. I think the chances of this camera containing a 64+ megapixel sensor are about the same as those of Rick Santorum leading a Gay Pride parade. But something in the 10–12 megapixel range would be reasonable, giving us somewhere around a 10:1 ratio of capture-to-final image pixels. This is of course educated guesswork, but capturing vastly more information per unit of image area is inherent to the way Light Field photography works. It’s all that extra data that bloats the file size (in megabytes) with respect to the image size (in pixels). Sensors are sure to get better, bigger and cheaper in the future, but conventional photography will always have the same ratio advantage over Light Field photography. Lytro cameras will be shooting 10-megapixel images only when (if) cheap point-and-shoot cameras get up to 100 megapixels.

General Disadvantages:
1.2 megapixel images
Images can’t be made into prints of reasonable size (this is really the same as the above: 1.2MP image size)
Images can only be viewed in special software
Images can’t be edited in industry-standard tools (Photoshop)
No control over shutter speed
No control over sensitivity (ISO) setting
No exposure compensation
Storage memory is built-in: you can’t swap out memory cards
Can’t have wide depth of field: you can choose a plane of focus but not to have everything in focus

Possible disadvantages:
Limited dynamic range – are Lytro images 8-bit color or 16-bit? No one seems to be telling but my money’s on 8-bit.

No raw capture… or only raw capture?
It depends on how you look at it. There’s no raw capture in the sense with which most photographers are familiar: non-demosaiced files in 16-bit color. But in another sense, all Lytro images are "raw format" because they’re in Lytro’s native, proprietary light field format. No JPEG but only a native file format that no third-party software can read. Worst of both worlds.

It’s hard to see the appeal to serious photographers who know how to manipulate depth of field. Or to point-and-shooters who generally want everything in focus. Still, there are some fascinating possibilities. The question is whether they’ll be of lasting interest or just novelties we become quickly bored with.

Other than the issues of raw capture and the ratio of data capture to image dimensions (and they are, admittedly, big ones), most of the problems can be mitigated to some extent as the technology improves. One major technological disadvantage will never be overcome and two big disadvantages have nothing to do with technology at all – they’re about human behavior. First, the technology…

#1 — Disadvantage to Everyone: Effect Doesn’t translate to Print

Obviously, as soon as you make a print the ability to choose the plane of focus is gone. Perhaps the spread of e-book readers may mitigate this problem… if e-book publishers and the makers of e-book readers decide to license the software for displaying Light Field images. Given the current pressure to create ever less-expensive devices, this is unlikely to happen in the near future and my money’s on "never". (There’s also the large increase in file-size vs. pixel dimensions in comparison to traditional images.) If you’re going to have to pick one focal plane before committing your image to print (or e-book) you’ll probably be planning far enough ahead to avoid the need for after-the-fact focus adjustment. Will there be exceptions? Probably. Enough to build a marketable product on? Probably not.

Now the human factors…

#2 — Disadvantage for Viewers:

This one is, in my opinion, the Achilles heel of light-field photography. Not only can the person viewing the photo choose the plane of focus, they have to choose a plane of focus. In other words, this technology makes the viewer do work. Remember Quicktime VR (Virtual Reality) 360° panoramas? Images that allowed the viewer to select their view in any direction from a central viewpoint? Cool at first, but then getting tiresome as every new image imposes a workload on the person viewing it. Imagine browsing a book of photography and having to choose the plane of focus on every image (or even just some of them). If the technology improves you’ll also get to (be forced to) choose the overall depth of field as well! Oh joy!

#3 — Disadvantage for Artists:

Here’s why I, as an artist, would never use this technology: The person who views photo gets to choose focus point, not the photographer who creates the photo. This goes contrary to the way most artists work: They want to be able to control the presentation of their work themselves, not have every person who views the work change it as they see fit. Will some artists take advantage of this ability in new and interesting ways? Probably. Interactive art has been made in various forms (like those 360-degree VR panoramas mentioned earlier) for decades now, but it has never constituted more than a small niche. It’s unlikely to ever be anywhere near a majority and choose-your-own-focus-plane photography will likely be a short-lived trend.

I don’t want viewers of my work second-guessing me about where the plane of focus should be!

So What’s it Good For?

As for good uses of this technology, I can see a few promising areas. First, as mentioned, there will be some artistic projects that effectively use the viewer’s participation and input as part of the work. This will be a small niche, but there’s no reason to think someone won’t be able to create really interesting art this way. Secondly, I see this technology as potentially useful in interactive displays in museums, trade shows, etc.

The big one I believe the creators of this technology are really aiming for is getting the technology embedded into phone cameras. They’d have a close tie between hardware, capture software and display software. Most people’s phone camera shots are shared largely on other phone cameras (via email, Facebook, etc), so the small pixel dimensions aren’t an issue. I’m guessing the real goal of the Lytro people has always been to have the technology bought by someone else – preferably a big phone maker. I still believe its marketability will be limited by the factors I’ve outlined above, but it’s one way they may be able to cash in on their invention.

There’s an oft-repeated phrase in the tech world: "You haven’t invented a product, you’ve invented a feature". That’s what the Lytro people have done. This doesn’t mean it doesn’t have useful applications or that they won’t find a way to make money on it (I hope they do, because it represents some truly creative thinking). It’s just not a revolution in itself – the way digital photography was when it took over from film.

Addendum, summer 2012

The Lytro revolution certainly hasn’t happened and seems to be petering out, if anything. In June their CEO announced he was resigning that position (though still staying with the company in an engineering capacity) in order to "focus on product vision". (To me this seems reminiscent of the politician who resigns his office in order to "spend more time with my family".)

Addendum, spring 2013

Developments seem to be playing out pretty much as I predicted: No noticeable impact in the general photography marketplace, no Lytro "revolution" and definite interest from cell phone makers. See my post on the subject.