»It should be a human right to be fairly represented by modern technology«
What is the technology behind how light and darker skin tones have been rendered and represented in film and photography? We talked to media professor Lorna Roth about the technological biases in cameras, about its history and challenges – and how furniture companies and chocolate manufacturers fit into that story.
[kon]: What does skin mean to you?
L.R.: Although I am fascinated with all aspects of skin including its anatomy and physiology, my work is focused mainly on the challenges we face when trying to represent skin tones in the visual industries, the arts, and intelligent design. To render flesh tones accurately or imaginatively in a final image product either analog or digital, we need to have an apparatus and sets of practices emergent from its design to be sensitive to a broad dynamic range of chromatic skin tones without having a dominant bias of one colour over another embedded in the materiality of analog film emulsion or digital camera algorithms.
I think of skin colour as a continuum of subtle changes in hues from the lightest of light to the darkest of dark and all of the incremental changes taking place from one to the other end of the colour spectrum. Therefore, much of my research has focused on technicalities, capacities, and visual industry practices and policies designed to render this range accurately within visual apparatuses and cultural practices.
I am interested in documenting visual industries’ incremental adjustments and shifts in the palette of skin colours selected to be fixed within the materiality of their image capture technologies. The objectives of my research are to reflect upon how these corporate design decisions are framed within historical and sociocultural contexts of the societies in which these colour adjustments have been rendered and, in turn, to assess the impacts of these on the social cognition of cross-cultural populations and diverse visual practices over time.
[kon]: In an article published in 2009, you describe the » Shirley Card«, a color reference point for lighting adjustments on cameras designed in the 1940s, originally depicting a white woman in a colorful dress, as »“emblematic of the state of race-relations/aesthetics in the industries of visual representations«. How has the »Shirley Norm« changed our way of seeing the world?
L.R.: Norm reference cards are important to the colour balance process in relation to analog and digital cameras as well as to lab printing. The process involves the determination of exposure, image density, comparison-measuring and calibrating the skin tone on photographs being printed. Historically, Kodak’s skin reference cards have almost always pictured a woman, and until the late nineties, one with light skin.
Shirley’s skin was considered the universal norm symbolically representing an ideal standard of beauty, tied to power, and privilege. She was often tagged »normal« either on the front or back of her image. Her light skin standard was chemically encoded within the materiality of analog film emulsion; it was also the preferred colour within digital cameras in their earlier iterations. Thus, »White Shirley«, a predesigned measuring template, became the »normal« as of the 1940s, leaving BIPoC as a deviation from the norm – living on the margins of the image world.
Given these design limitations, I don’t think that the initial Shirley norm reference cards changed our way of seeing the world. What they did was deeply entrench within visual technologies and industry practices an already existing unconscious way of seeing people with dark complexions. To be more specific, it enabled light-skinned people to become significantly more visible on film and in photographic practices than those with darker skins. On the contrary, the same technology and industry practices seemed to depict the latter in ways that obscured their features, making them somewhat invisible and less central to the photo. There was an obvious, though unconscious, racial impact of this decision on the practices and protocols of imaging human subjects over the years. Even in the nineties, when three women of differing racial backgrounds began to appear on reference cards, they were referred to as multiracial »Shirley’ Cards«, nostalgically reminding the industrial lab workers of a time long passed.
Take a look at how Shirley Cards work.
[kon]: How and when did photo labs became aware of this problem?
L.R.: After much consumer feedback during the desegregation process in the United States, photo lab workers became aware that rendering of the details of dark skin tones was highly deficient in its absence of a technique and a set of tools for representing the full spectrum of skin colours of the North American and international populations. They realized the necessity for the development of compensatory measures and technology improvements to redress its shortcomings, mainly, its limited dynamic range. To compensate for their films’ deficits, they recommended using stronger lighting, faster film speeds, make-up appropriate for dark skins, and larger aperture openings. These eventually became common practices among custom photographers, mostly from the African-American communities, who had figured out, before stock film producers, how to modify and create photos of subjects with darker skins that appeared very beautiful.
Several factors drove Kodak to make changes in the chemistry of their film emulsions and later their Shirley cards. These included Complaints from parents of racially integrated high school students in Rochester, New York in 1959, that end-of-year photos were off-putting. Details of Black students’ faces showed very poorly, if at all, while the whites of their eyes and teeth stood out. Meanwhile, lighter-skinned students’ images showed much greater detail while appearing to be overexposed. Although Kodak was concerned about these criticisms from the children’s parents, they began to consider how to deal with the problem, but did not find a solution until the seventies.
In 1974, Kodak received complaints from wood furniture and chocolate manufacturers about the lack of dynamic range for browns on ads they produced for clients. It was too hard to differentiate between dark and milk chocolate and dark and light wood grains. When Kodak added more shades of brown to their chemistry, the incidental consequence resulted in much better rendition of the fine features on faces with darker skin tones, not easily seen in prior stock films. This made a big difference – theoretically, but not yet in practical action.
»The designer or modifier of a visual technology is always apparent in his or her design.«
It actually took two decades before they publicly responded. In 1996-97 Richard Wien and his team at Kodak, produced the first multiracial Shirley Card showing a Black, Asian, and Caucasian woman together, still using the name Shirley as an echo of its origin.
In viewing the cards widely used around the world but particularly in the US context, one can almost trace a linear relationship between the Shirley images circulating and the civil rights movement, desegregation, integration, immigration, multiculturalism, and multiracialism. The designer or modifier of a visual technology is always apparent in his or her design. His or her values and biases can be discerned by carefully studying the chemistry of analog film, and the algorithms embedded in the materiality of the digital apparatus.
[kon]: In the same article you also describe a reference image by Getty Images depicting women with a wide skin color range along with a kangaroo for greyscale and colorful household objects as an excellent example of »sociocultural inclusivity«. This image has since been taken down from Getty Images’ Servers. Do you think they decided it was no longer of interest to them?
Image commonly used for calibration verification. Located within the Getty Collection of photos on the Internet, 2009. (Printed with permission of Dry Creek Photo)
L.R.: I still consider this image which I originally sourced from Dry Creek Photo in 2009 to be one of the most inclusive and interesting reference cards used for the colour balance calibration process. In pondering the global history of skin colour adjustment standards, the gradual expansion from the traditional single Shirley image to the more inclusive range of skin tones, animal furs, colour bars, grey scale, and a rainbow of object colours, seemed to provide a much more realistic representation of international populations and colour samples. In other words, its dynamic range was much wider than former single-person cards that preceded it. Consequently, I’m not at all certain why Getty would have chosen to remove it from circulation within their collection of images. However, I can suggest a few possible reasons.
As identity politics became more popular around the world, national, regional, and local digital lab leaders might have considered that Shirley cards would be more appropriate, easier to create in each individual lab, and more cost effective if they reflected the actual multiracial populations and nations within which their businesses were situated. Another possibility might have been the risk of the Getty card becoming reified as another or a new single standard within a cognitive framework centered on or limited to the examples of diversity as embedded in this particular card, still leaving out multiple constituency groups each with differing skin tones.
Finally, we should not forget that these cards were initially very expensive to create and distribute around the world. Even the single Shirley card had become the subject of lab complaints about its high cost. Consequently, individual labs which created their own cards reflecting their local cultures and populations were likely to be seen as more relevant to the visual industries in their home regions, would cost less to produce and distribute, and would relate more clearly to the identities of the population with whom they would be working. Thus, there was less demand for a pre-fixed, singular norm, and more interest in a flexible card design based on local cultural reference images.
[kon]: Why does it matter to not just look at the systemic bias in photography but also to focus on the technology behind it itself?
L.R.: Until very recently, most of the criticism about photography has been focused on image deficits that are evident in visible representational practices such as absences of racial, ethnic, gender and other minorities as image subjects. At a deeper level, these absences to some degree are tied to a technical apparatus dysconsciously designed to leave out easy rendition of darker skin tones.
We can no longer consider visual apparatuses to be politically innocent or neutral. By now, it is well known that light skin colour had been encoded within earlier versions of stock film emulsions and tonal range in digital cameras. This is no longer a contentious comment.
»Before it became known to black communities that the reason their home photos did not capture the fine details of their faces was not their fault, they suspected that they were incompetent photographers because they lacked training and skill.«
As I’ve said before: Beginning in the 1950s, consumers’ and prosumers’ critical feedback about the too narrow dynamic range of skin colour representation had driven companies like Kodak and Fuji to make colour adjustments. However, these were mostly motivated by economics, rather than constituency group lobbying protests, which was not very surprising. At the time, constituency groups were more preoccupied with economic concerns and civil rights rather than aesthetics. Cameras were mostly used to record family histories and most people assumed that the quality of images was the outcome of a purely scientific process based on physics of light.
Before it became known to the black communities around the world that the reason their home photos did not capture the fine details of their faces was not their fault, they suspected that they were incompetent photographers because they lacked training and skill. However, the foundational root of their poor detail rendition was in fact because both the analog film emulsions and early digital cameras they were using were not designed with darker skin in mind. Black people have had to either innovate and experiment with compensatory strategies themselves to improve the quality of skin tones or wait until enough criticism had made financial impact on the industry for their business executives to invest in research and development targetting more inclusive products.
It should be a human right to be represented by modern technology equitably and in as fair a manner as possible. Think of it as a form of aesthetic justice: racial equity which can take the form of a broad continuum of colours in the spectrum, built directly into the technology, and reflected in industry policies and external photographic practices that match the inner workable possibilities of the apparatus in use. It matters to look at both the systemic bias in the end product, the photo, as well as the actual algorithms of the technological infrastructure on which the final creative image is dependent.
[kɔn]: In January, Vanessa Nakate, a Ugandan climate activist, was cropped out of a photo issued by the Associated Press. The original photo shows her among other female climate activists that attended the World Economic Forum in Davos. AP later apologized, stating »compository reasons« were the motive behind it. However, a side by side comparison of the two photos shows that the White skin tones of the other women in the photo are much brighter than in the one that was cropped. Considering, that representation is a set of practices driven (amongst other things) by technology: What do you think are the decisions a photographer in this situation unconsciously makes when she takes a look at his or her photo that technology shaped?
L.R.: I suspect the photographer’s decision to eliminate Nakate from the photo of five young women, including Greta Thunberg, had as much to do with I don’t think the photographer’s decision to eliminate Nakate from the photo of five young women had much to do with her skin colour and her skin colour as it did with more to do with his focus on the »look of the image«. In an article in The Guardian, he was noted as having given his rationale to be that »the building in the background was distracting«. However, without Nakate, the woman to the far left is situated quite close to another building that could also be considered distracting. Why wasn’t it a compositional problem for this second person? In the primary photo, there was a short woman’s face hovering over Nakate’s left shoulder and, in my opinion, the white spot above her shoulder, which was a portion of another building, was not offensive.
The photographer’s rushed compository decision created a negative impact which distracted the viewer from the essence of what should have been the central message that I think he was trying to make about the conference. As a matter of fact, it altered and distorted the focus of the story by eliminating the presence and voice of the Ugandan representative whose country and continent are highly affected by the climate crisis. Instead, he chose to display conference delegates with light complexions only, inferring that the activists who are working to promote solutions to the global climate crisis are all of European or Western (light-skinned) descent. In doing this, he missed the discourse that had materialized from the international conference, as well as the cross-cultural alliances that participants had the opportunity to develop for future policy strategies.
As I am an educator, I would think that this issue could make an excellent basis for a public debate about why an international discussion about such an important topic as climate change ends up being refocused on the elimination of the only black participant in the photo under the guise of »aesthetic« reasons. A discussion of such invisibilization in Schools of Journalism around the globe would certainly have pedagogical value.
[kon]: The ongoing Black Lives Matter protests in the United States are highly sensitive issues in regard to visual representation. It is striking, that very rarely pictures taken at these demonstrations show people with White and Non-White skin tones together in one frame. What impact did photo companies have on visual representation of civil movements now and in the past?
L.R.: During the period of black and white segregation in the United States, well before colour balance accommodations were considered to be a challenge – up till the late 1960’s –, there were very few public opportunities for Afro-Americans and lighter skinned people to be in the same photo because few occasions attracted both groups to be together in the same public space. As desegregation evolved, it slowly became a more realistic possibility. However, results were not flattering technically; thus not very encouraging in terms of photo aesthetics.
Two children in Senegal, Natalie Le Brun (Right) and Guilado Sarr (or her sister) (Left). Circa 1973. Kodak film (125 ASA) used with Canon camera. An example of the challenge of photographing highly contrasted skin colours in the same frame and the difficulty of later recognizing who is in the photo. (Courtesy of Olivier Le Brun, Paris, France)
The changes in the Shirley cards, in the stock film emulsions and in the critique and reformulation of the algorithms which controlled the width of the dynamic range of various cameras, thus came into being in the context of sociocultural and political changes catalyzed by the civil rights movement. In fact, what occurred from the mid-seventies to the mid-nineties was the beginning of the desegregation of visual technologies which slowly took shape in tandem with the successes of the civil rights protesters. These societal transformations led to the opening of a new ideological and pedagogical framework for thinking and teaching about image capture.
With this as background, I want to turn to journalism training in the US and elsewhere. For a very long time, »objectivity« reigned supreme in news media coverage and was equated with the notion of »two sides to every story«. Thinking in such a dualistic manner doesn’t leave much room for or enable the co-presence of more than two perspectives at a time.
»We can hopefully look forward to improvements in the quality of pictures in which whites and non-whites will be sharing a single screen.«
The coverage of demonstrations and protests, related to Black Lives Matter, racial profiling, anti-racist gatherings and many other crowded public assemblies where participants are of different racial origins continue to be depicted in ways that separate activists that have been victims of brutality from those that are there as allies in supportive positions. It’s often assumed that alternative media journalists will cover a wider range of opinion of those present who are not central to the leadership or core members of the team that organized the protest.
To further complicate the issue, one has to take into account the skin tone/race of the journalist in terms of ease of visible access into and out of the occupied space of a protest against racism as journalists are not always welcomed witnesses. Given that there are more diverse supporters and allies for civil and human rights issues in the 2020’s, and given that most photo companies have shown their good will to correct the colour infrastructure of their visual technologies, we can hopefully look forward to improvements in the quality of pictures in which whites and non-whites will be sharing a single screen.
