It was only a matter of time before an increasingly number of computer scientists began wrapped their heads around digital imaging in a big way, at least in their spare time. That's exactly what Carlo Baldassi, a student in computational neuroscience did, after looking at some pictures of his girlfriend that appeared too constrained and out of proportion. Baldassi has created an automatic photo-editing software tool that always the user to stretch an image without it looking stretched. Peter Wayner's article in The New York Times quotes Baldassi as saying, "Reality is a lie." Nice quote perhaps, but the implications are much more far-reaching as software such as the one Baldassi has made becomes commonplace.
Automated tools like Mr. Baldassi’s are changing the editing of photography by making it possible for anyone to tweak a picture, delete unwanted items or even combine the best aspects of several similar pictures into one.
The tools are giving everyone the ability of the Stalin-era propagandists, who edited the photographic record of history by deleting people who fell out of favor.
Wayner's last statement is a bit troubling. Sure, we have the tools now to seamlessly stretch the truth, but do we need to? In my on-going survey on digital manipulation more than 40 percent of respondents indicated that they could tell when a picture had been altered.
2007-2008 snapshot of the photo manipulation survey related to whether people can tell if a picture has been altered.
2006-2007: Note that the sample sizes differ considerably.
During my time surveying people about digital photo manipulation, a fairly high percentage of people report they can tell when a picture has been altered. I find this opinion interesting, because in my own experience I am not as skillful.
In my own experience, I find myself having less time to carefully scrutinize pictures. I do assume, though, that there is an increase in altered images in the media with the introduction of digital technologies, but because of the volume of pictures flooding our consciousness, I tend, like many people, to just scan images quickly. I tend to judge the authenticity of a picture on the context and source in which it is disseminated. For example, I would tend to trust the authority of a news image in The New York Times over an advertising image any day. This means that I wouldn't typically spend time looking for manipulated images in The New York Times, while I just assume that most advertising images have been altered to varying degrees.
Getting back to Baldassi's software, which is based on the seam carving work of Shai Avidan and Ariel Shamir, it makes sense that many of these tools will become commonly accepted by people over time. In the future, we will just expect that the images we see have been enhanced in some way and that the notion of objective reality is nothing more than a passing fancy.