I first heard about eye-tracking technology in relation to my work for a consumer products company. My company was using it to design shelf space in stores. Of course I was straying off on tangents and thought that it would be interesting to use the technology to see how people look at photographs. Well, people are looking into it.
Cool study. Could this explain why we concentrate so hard on ensuring all the elements of our images are perfect and our non photographer viewers seem to ignore obvious flaws in the image away from the main subject? I often have had friends or co workers rave over an image I thought was second rate. Maybe we now know why?
Posted by John W. Wall on March 19, 2007 at 7:00 PM
Ed's question reminds me of a bit of training that we do as photographers. How many times are we sure we got the shot, only to see later than some errant twig or whatever was in the shot? Or the old problem of having a tree growing out of someone's head? So we train ourselves to look not just at the main interest, but the other stuff as well. I also wonder if some of what we think of as second-rate characteristics -- the ones ignored by friends and coworkers -- are partly a matter of living *inside* the box of what's "supposed" to be good.
I wasn't interested in this work so much because of what artists look at, but, rather, because of what normal people look at. Well, maybe it was the discrepency between the two. Is the box we're working inside the wrong box? Is the box ok? Could you use eye tracking to a box maker's advantage?
Pete D.
Posted by Lon Overacker on March 20, 2007 at 9:27 PM
Pete,
Quite an interesting story, thanks for posting it.
For me this just confirms a couple of things regarding composition. We as photographer's critique and nitpick our images from an "artists" point of view; examing the entire image, finding those twigs at the edge of the frame, etc. etc. But if we back up, step out of the box as Pete suggests, then the old cliches become actually valid and important. Things like the "rule of thirds," or don't split your horizon down the middle or making sure your main subject is in focus. It's also why non-artists might not enjoy some of our favorite "chaotic" images of nature - there's no visual focal point. If you're shooting with an audience in mind, perhaps selling a print, then we should be conscious of what is appealing to the non-artist.
I think also, as you may be suggesting, that we step out of the box in critiques of our own and our peer's work?
Lon
Posted by Troy Gizzi on March 21, 2007 at 12:47 AM
Interesting article, Pete. Thanks for sharing that. I'm a software developer and it may interest you to know that eye tracking is also used to evaluate the usability of software interfaces (such as web pages and Windows forms). I had no idea it was being used elsewhere. Pretty cool stuff.
Discussion
6 repliesCool
Pete D.
I wasn't interested in this work so much because of what artists look at, but, rather, because of what normal people look at. Well, maybe it was the discrepency between the two. Is the box we're working inside the wrong box? Is the box ok? Could you use eye tracking to a box maker's advantage?
Pete D.
Quite an interesting story, thanks for posting it.
For me this just confirms a couple of things regarding composition. We as photographer's critique and nitpick our images from an "artists" point of view; examing the entire image, finding those twigs at the edge of the frame, etc. etc. But if we back up, step out of the box as Pete suggests, then the old cliches become actually valid and important. Things like the "rule of thirds," or don't split your horizon down the middle or making sure your main subject is in focus. It's also why non-artists might not enjoy some of our favorite "chaotic" images of nature - there's no visual focal point. If you're shooting with an audience in mind, perhaps selling a print, then we should be conscious of what is appealing to the non-artist.
I think also, as you may be suggesting, that we step out of the box in critiques of our own and our peer's work?
Lon