AKQA IMPACT / 2023
Historically, film technology wasn’t developed to capture all skin tones. For many years, white skin was used as the calibration standard, often through “Shirley Cards”—images featuring women with light skin tones, the first of whom was the eponymous Shirley. This made it challenging to accurately portray Black skin, with images appearing blurred, flat, or poorly shaded. Such biases persist in technological advancements today. RGBlack is a social impact platform designed to spotlight photography’s unseen bias.
↳ role: concept/ creative idea / copy
rgblack
↴
The legacy of the Shirley Card
The original Shirley Card featured an image of a white woman with reference points for color, exposure, and density. This quickly became the industry standard for calibrating printing technology, and photographers relied on these cards for years to establish what was considered "normal" in photo processing. In 1990, multiracial Shirley Cards were introduced to accommodate a broader range of skin tones. However, these new cards continued to depict lighter complexions and were never widely adopted (as they coincided with the emergence of digital photography).
↴
Recalibrating inclusive representation
The rgblack.org platform introduces a new set of calibration cards specifically designed for diverse skin tones, alongside resources on lighting techniques, aesthetics, and color science. By recreating the original Shirley Cards, RGBlack imagined a new approach to photography—one that celebrates inclusivity.
To mark the launch, photographer and director Juh Almeida of Pródigo Films created a film that reinterprets Shirley Cards through a Black perspective. This film encourages reflects on representation, both behind and in front of the camera, and explores the creation of the new Shirley Cards.
Media coverage of RGBlack’s launch has sparked conversations around racial bias embedded in audiovisual technologies and culture, underscoring the urgent need for diverse perspectives in the industry.
↴
Addressing the bias in AI
Today’s AI image-editing tools are typically trained on mass digital image libraries. Often, these carry inherent bias. Studies reveal that identifying and removing these biases is complex, resulting in AI frequently misrepresenting or excluding real people—repeating the limitations once reinforced by Shirley Cards.
