Explainer

Were the Dove Ads Retouched?

How to spot Photoshop chicanery.

A Dove “Pro Age” advertisement

After a New Yorker profile implied that “king of the photo touchup” Pascal Dangin had airbrushed photos taken by Annie Leibovitz for Dove’s high-profile “Campaign for Real Beauty,” the company issued a statement last Friday explaining that Dangin had only removed dust and performed minor color corrections. Is it possible to determine whether the Dove photos were retouched?

Maybe, but it would be very difficult. Amateur retouching can leave seams where two different images are spliced together. But in the case of an expert retoucher like Dangin, visible signs would have been diligently scrubbed away. Additionally, since images can be distorted when they are compressed into other file formats (PDF) or printed in a magazine, any apparent smudges or irregularities in one of the Dove photos might well be artifacts of the photo’s reproduction, rather than signs of tampering.

A more advanced form of analysis might focus on the lighting in the photos to see if there were inconsistencies in the images’ shadows or lighting gradients. (See, for example, this forged photo of John Kerry and Jane Fonda, where one calculation (PDF) pegged the light coming from 123 degrees onto Kerry and 86 degrees onto Fonda.) Another technique used to spot composite photographs looks at the way light reflects off a subject’s eye to determine exactly where the light was coming from. Professional retouchers say they can also recognize where a photo has been altered if the skin texture is too smooth or the model’s symmetry too perfect, or if the perspective appears distorted.

But even the most advanced analysis of the Dove ads might not reveal very much. Because the photos were taken inside a studio, Leibovitz and Dangin didn’t have to worry about the complications that come from natural sunlight. The setup of the Dove photos was also rather simple—just the model alone—so a skilled retoucher would probably be able to ensure that any corrections kept the lighting consistent.

If an image-forensics specialist could get a copy of the actual image used in the ad, then he or she might be able to conduct a more detailed analysis of the digital information that encodes the image. Had Leibovitz been using a digital camera—until recently, she used only film—then analysts might be able to use a technique (PDF) pioneered at Dartmouth College that takes advantage of the way digital cameras create an image when photos are taken. Nearly all cameras directly record just a fraction of the pixels that make up the photograph and then “interpolate” the rest based on patterns within the image. As a result, any uncorrected section of a photograph should fit these correlations, and a computer analysis can show if the photo has been altered even slightly. Of course, since Dangin and Leibovitz have already admitted that the photo did include the minor dust and color corrections, this method wouldn’t be so helpful.

Got a question about today’s news? Ask the Explainer.

Explainer thanks Brian Dilg of Brian Dilg Photography, Katrin Eismann, Hany Farid of Dartmouth College, M. Kimo Johnson of the Massachusetts Institute of Technology, and George Reis of Imaging Forensics.