By now you’ve seen the YouTube video of two very observant retail store workers, Desi and Wanda, that “proved” that HP webcams cannot process facial recognition of African-American skin in low-lighting and without optimizing settings for each skin tone. The Webosphere is gleefully hurling accusations of racism at Hewlett-Packard that show a woeful lack of knowledge from those accusers about how imaging technology works.
As an African-American with a darker skin hue, I can attest to many casually taken smartphone pics of me and groups of friends (usually in some dimly lighted drinking establishment), in which my brown skin shows up in images as an inky blur, blending into the background. In contrast, pictures of Caucasian friends often result in glaring milk-white skin, accentuated by the flash.
Same thing with webcams. More light reflects from light objects than from darker objects. There is no conspiracy, no racism on the part of Hewlett-Packard: It’s Science 101. In fact, you can easily find lots of articles from professional photographers about taking digital images of African American skin. Here’s one from the New York Institute of Photography’s site:
No one has a clearer explanation than our good friend and master portrait photographer Monte Zucker. As NYI students learn in Monte’s portraiture videotape that is included in Unit Six of the Complete Course in Professional Photography, it’s simple.
“If I am lighting a black person, I’m not going to change the light, I’m not going to bring it in any closer. I’m not going to open up an extra f-stop.
“The only thing I’m going to do is use the light coming from the side and around the subject. What we need to do when we’re photographing a black person is to bring an extra light in from a 90-degree angle.”
Accusations of racism in this case are as nonsensical as saying it’s discriminatory that blue eyes turn up in flash photos as red much more frequently than darker eyes. Imaging technology is about precision, pixels, resolution, and lighting; not race judgment, thank goodness. Judging still remains a uniquely human trait.