On May 31st, 2017, the sun having set on a day that ended with a series of heavy thunderstorms, I took a photograph with my smartphone out of the window of a diner. It was the kind of imagery impossible to resist: the golden glow of the fading day multiplied with ominous low-hanging clouds in the sky, the light reflecting off the wet ground, with the diner's neon lighted sign blending in seamlessly with a sea of yellow and orange and magenta tones. Perfect Americana kitsch, however real it actually was (Figure 1). I posted the image on Instagram, anticipating a flurry of “likes,” which indeed trickled in. There was no need to use any of the app's filters, reality had already provided it.
On each of the different devices I own – a desktop computer, a laptop, an iPad, and my smartphone – the picture looks a little different. This might sound counterintuitive. But each of these devices has a different type of display, rendering colours and details differently (on top of that, my phone sports a “hot pixel”). There are, in effect, four versions of my one picture in my own home, and there are countless others out in the world – wherever someone comes across one on Instagram.
Once a photograph has existed long enough on the internet, it spawns even more variations, caused not just by variations in display hardware, but also because of software issues. A simple Google image search can demonstrate the effect. For example, on May 28th, 2016, I used the search terms “Eggleston Red Ceiling,” which returned a grid of the well-known photograph, or rather a grid of versions of the image. In Photoshop, I selected each image and had the software produce its colour average (Figure 2). There is a noticeable difference in the results. One would naively imagine that all thumbnails of the very same photograph have the same average colour, but they don't. The source of these variations is not immediately obvious. I'm guessing it's a basic question of colour spaces, of how many (and which) colours were rendered and/or stored when each new copy of the original photograph was created (note that this assumes that there only was one original copy, which might not be the case at all).
Another, more well-known problem concerning photographs online is created by compression algorithms, which usually means the use of the jpeg format. Such algorithms reduce the amount of data, while, when used smartly, only throwing away non-visible information. But of course, more often than not, the resulting image degrades, an effect that is also well known for video. Currently, many discussions around the medium of photograph revolve around its veracity, especially in this age of “fake news.” However, for photographs and videos, the opposite of what one might expect seems to be the case. In light of “citizen-journalist” images and especially videos being so widely disseminated, obviously blurry or visually distorted images or videos are not at all seen as less credible. In fact, it's usually the professionals' images that face increased scrutiny, especially in cases where they seem to look “too good” (a good case is provided by World Press Photo 2013 winner Paul Hansen, whose image looked manipulated – essentially too good – for a lot of people1).
In the days of analog photography, materials-based variations tended to be indirectly built into the machinery used to make pictures. If you owned a Polaroid camera, for example, you knew that there could – and would – be variations in the outcomes. As I outlined above, in the world of digital photography, there also is ample room for error, but it manifests itself in different ways, not the ones we want. The world of especially smartphone photography is now filled with ways to artificially create the “look” of analog photography. It turns out what previously was seen as irritating – unpredictable materials – is what people want, even just in simulated form. Polaroid gave up on making film, but it now sells an app called “Polamatic,” which re-creates the look of various of the company's previous films. The company achieved this result by scanning old pictures. For example, there are 12 SX70 film filters, or 23 distinct (white) borders. Of course, this falls way short of what an SX70 user would experience in the past, where each new film pack (and camera) would be a little bit different. Instagram filters operate along the same lines of thought. There now are even apps (such as Oblique) that create the kinds of image distortions rogue digital algorithms might produce – glitches.
Theoretical or critical thinking around photography ought to consider the popularity of smartphone photography filters or apps that simulate older materials. For too long writers such as Susan Sontag have had too much influence with their dismissal of how photography was used by the uncouth masses (this wasn't how Sontag would phrase it, but between the lines, her disdain is quite obvious). There now exists a real demand for visuals that the world of analog photography tended to treat as nuisances. Crucially, such “defects,” whether real in the analog and simulated and/or real in the digital world, are usually not seen as damaging the credibility of images. In fact, they instead seem to have become a part of photography. Now that camera makers have spent so much time on making them disappear, people want them back. This has got to mean something.
In addition, and this is far beyond the scope of this essay, it might be time to re-consider our ideas what photographs really are. In light of the many small, yet noticeable variations in the photographs we produce, we might have to move beyond the idea that as mechanically or digitally mass-producible entities they are perfect clones of each other. The vast majority of photographs now lack all physical form. They exist somewhere as “data,” to be displayed, however briefly, on some screen, in the form of an imperfect clone. In light of the preceding, it seems clear to me that collectively, we prefer our photographs to be imperfect.
But if any of the pictures I'm looking at really is just an imperfect clone, possibly in a line of past and future clones, what exactly is the picture itself, the original picture – or however else one might want to describe it? Coming back to my diner photograph, I'm currently left with the thought that the original image doesn't exist in any other form than code, the very code that my various devices translate into the slightly different clones. In the analog world, for those images that were cloneable the code would have been the negative.
And these latent proto-images, translated as they were or are into photographs, subject to machine, software-, or user-based imperfections, then meet our expectations of what a photograph ought to do and to look like. It might be here, at the level of our expectations, that we might want to consider the role or importance of imperfections or defects, of why and how pictures look the way they look.