Is CGI Software Poised to Kill Photography? It's Close

Despite public craving for ridiculous Photoshop failures and public shaming of horrific photographers, the perpetual myth that a fancy-ass camera and technical wizardry is the way to get rich off photos still remains. Outside of the hyper-competitive world of fashion shooters and international photojournalists — and the decidedly un-glamorous world of city news, which finds photogs shooting sidewalk blood stains on an uncomfortably regular basis — the path to working success for a young photographer tends to involve grinding out a thick portfolio of stock images and product shots.

Stock photos may not be super exciting, but they certainly help pay the bills. They’re also becoming increasingly easy to replace. Don’t believe me? Take a look at the next two images and tell me which one’s fake.

Sorry, you’re wrong. They’re both fake. Photographer Mark Meyer whipped them up with a cheap bit of CGI software, and according to him, it’s easy as pie. In an excellent post on his blog, Meyer notes that while photography was long the cheapest way to capture realism, that’s changing. Thus, as software becomes increasingly powerful and cheap, photography increasingly looks like a dated technology. He writes:

Computer generated imagery and photography are on intersecting trajectories. While photographers employ tools like Portrait Professional that sanitize their portraits, making them look more like renderings, 3D artists are adding blemishes and developing tools like subsurface scattering to make their renderings look more like snapshots. Photographers are fighting to remove noise, CGI artists are adding it; photographers are using digital techniques like focus stacking to extend depth of field, while CGI artists begin with unlimited depth of field and artificially reduce it. At the moment photography is still the most affordable means to quickly create realism in most applications with notable exceptions in large scale cinema productions and car advertising. But the two worlds are about to merge and a large part of the photography industry will be replaced by software.

Now, before we go any farther and people start freaking out, let me remind you that we’re talking about photos as products, not as art. When it comes to an ad for a watch or a gadget, firms don’t care if a photog shot on a Hasselblad or a Mamiya, they just want smart, clean images. If a computer can do the same, with more flexibility (changing perspective after the fact, for one, is a huge potential selling point), why not?

It’s not like this is the first time photographers have worried about being replaced, either. Wander into any photo forum online and you’ll find a wealth of people complaining about how cheap, easy-to-use DSLRs and a wealth of online photo-publishing platforms have flooded the world with a ridiculous wealth of images, ranging from great to heinous.

Now, I’d tend to view those guys as a bunch of crotchety complainers, like many forum people tend to be, especially when you have guys praising examples of poorly-designed complexity in modern cameras because it makes it harder for non-pros to use them. But when I’ve found myself wandering through a favela in São Paulo only to come across a family stumbling about, waving $10,000 worth of gear around for aimless snapshots, I guess I understand a little bit.

Then there are more egregious examples, like that of David Shankbone, who’s allegedly been called "arguably the most influential new media photojournalist in the world." Why’s that? Well, he’s taken the photos that have been used in over 5,000 Wikipedia articles, thanks to the fact that he licenses all of his images under a free Creative Commons license.

Paparazzi make a ton of money, and as such are a rather competitive bunch, and they surely don’t take kindly to Shankbone giving away all his work for free. But Shankbone’s hardly the problem; for every one of him, there are thousands of other people putting their shots out their to be seen, rated, and distributed for free.

From a publishing standpoint, that’s great. Motherboard would be a hell of a lot more costly to run if we had to pay licensing fees for every image we use. And from the standpoint of an amateur or someone who shoots a lot on their own time, the ease of spreading a portfolio online means getting exposure (and finding good photos) is more egalitarian than its ever been before. But, once again, if photos pay the rent, the huge online photo community doesn’t help.

But what if CGI and the mountain of images on the web were combined? Well, that’s already happening. In a paper published on the Association for Computing Machinery’s Queue Blog, David Crandall and Noah Snavely discuss how to apply the methods of computational photography to the vast wealth of images online to not only get a better spectrum of images of a single place — see geotagging on Flickr — but to also create whole new images. Let me borrow a slide from their post to illustrate what I’m talking about:

Thousands and thousands of photos of the Colosseum are easily available online, but the best are most likely going to cost money or involve some sort of licensing. Look at the image Crandall and Snavely produced: It’s well-composed, relatively distortion-free shot that would normally need the right equipment, skill, and potentially the stitching of a number of separate images by a pro. Crandall and Snavely skipped all that:

The reconstructed cameras are shown as black wireframe pyramids indicating where each photo was taken, and the Colosseum is reconstructed as a dense 3D point cloud, similar to what a laser scanner would capture—but in this case, reconstructed completely automatically from photos found on the Internet.

Just think about that for a second. Here’s a photo with a custom perspective and lighting, developed totally from images ripped off the Internet. Sounds like a dream doesn’t it? I mean, why would anyone pay for a photographer when they could take the best parts of a bunch of random Internet pictures to make their own perfect shot? No wonder people are saying that photography is screwed. The only problem is they’re wrong. As Meyer writes:

For the first time in history, photography is about to lose control of its monopoly on affordable, convincing realism and it’s time for us to understand that realism has never been the most important feature of the photograph. Although we rarely think about it, we understand this intuitively: a computer rendering of your daughter’s wedding will never be the same as a photograph even if both are equally realistic. The photograph is defined by its causal, mechanical connection to the real world.

Look at it this way: Photojournalism will always persist. Look at award-winning photographs of powerful moments, and then try to tell me that they could have been replicated by computer processing of other shots (impossible) or by some dude wandering by with a cell phone (nearly impossible). Good photojournalists will always have jobs because it’s impossible to replicate what they do. (Shrinking photo budgets are of obvious concern, but aren’t directly the fault of technology.)

And, although photojournalism is a particularly stable example, that thread carries through to anyone who takes pictures for a living. So are photographers poised to go out of business? Well, people grinding out product shots for cheap employers are probably going to feel the pinch, but as for anyone else, people simply don’t want a substitute.

People demand the real deal, and aside from the aesthetic advantage of a good shooter versus a computer rendering, there’s the simple fact that people will always have a better connection with an image that was captured by someone who was actually a part of that moment. It’s as simple as that: You feel disappointed as soon as you real an image isn’t real, because capturing reality — not realism — is the whole point of taking pictures.

Follow Derek Mead on Twitter: @drderekmead.

• Obsolete: Photoshop Before Photoshop
• How Hollywood Abuses the Color Wheel
• The Lolcats of the 19th Century