Kodak, Racial Bias and the "Shirley" Cards

Moderator Note. This has gotten excessive, and basically into a conflict of belief systems and news sources - IMO the piece was political as are many responses. Please leave it or I will start deleting.

Films were made to allow us to capture images of the world we all share, and improved across time. I recall difficulties accurately capturing scenes with wide ranges of values, regardless of the film. Back in the day there were discussions about variations in films renderings based on where they were produced - Agfachrome, Fujichrome, Kodachrome and having advantages or disadvantages - then there was Ektachrome...

Moderator Note. This has gotten excessive, and basically into a conflict of belief systems and news sources - IMO the piece was political as are many responses. Please leave it or I will start deleting.

 

Films were made to allow us to capture images of the world we all share, and improved across time. I recall difficulties accurately capturing scenes with wide ranges of values, regardless of the film. Back in the day there were discussions about variations in films renderings based on where they were produced - Agfachrome, Fujichrome, Kodachrome and having advantages or disadvantages - then there was Ektachrome...

Sandy, I apologize for my side of the "dialogue" between the other party and myself. I allowed myself to be drawn in when I read his early post.

Maybe the technician should use a grey card or other tech to get the right tonal settings and color temperature. Was it not digital video? Maybe the tech didn't know what she was doing.

I'm not very familiar with the standard methods of settings up video cameras so I didn't want to say too much about this. But I'm pretty confident that the video tech was not real knowledgeable about it.

My expertise is more in the realm of photographing and printing with professional (portrait/wedding) color neg films. I moved into lab work in perhaps the mid to late '70s as a way to learn about the (then) "black art" of color photography. I found this sort of work more interesting (and challenging) than pure shooting (I had been doing moderate amounts of weddings, and plenty of high volume portrait shooting). In the lab case I eventually became the QC dept manager, so I like to imagine that I have some knowledge in that area.

At any rate I find the author of the NYT article to be, uh..., somewhat "lacking" with respect to knowledge about photo lab operation. I'll give a few examples, their quote, then my [A]nswer:

[NY Times:]For example, developing color-film technology initially required what was called a Shirley card.

[A.] No it didn't. A lab could function perfectly well without the so-called "Shirley card." Note: I don't know where they get the "card" terminology; it doesn't come from the Kodak catalog. In the labs I've been involved with they're seldom even called Shirleys - they tend to call em "slope negs," "master negs," or the more proper catalog term, "Printer Control Negative Set."

[NYT article] ...lab technicians would use the image of a white woman with brown hair named Shirley as the measuring stick against which they calibrated the colors. Quality control meant ensuring that Shirley’s face looked good.

[A] This is news to me. None of my QC people were look at the "Shirley" face to see if it looked good. For "process (QC) control" we used densitometers to read "control strips" which were plotted on charts. For "product quality" I had a full time inspector who would semi-randomly go through samples of work ready to ship. One of the inspection items was, "are the prints within the aim spec (5 cc units of color for us)?" Now we used our own standards for color, with reference images of several complexion types, printed in a "ring-around" to show the acceptable limits in different colors. In a small scale operation you don't need to do this, but with multiple people in different areas it's difficult to function without reference standards. But a pro lab should be making their own standards.

[NYT article] In the mid-1990s, Kodak created a multiracial Shirley Card ... in an attempt intended to help camera operators calibrate skin tones.

I have no idea what it means for a "camera operator to calibrate skin tones." I'm skeptical that the author knows either; I would be interested in hearing their explanation. (Perhaps the article is jumping back to video cams?)

The author seems to put a lot of weight in comments by a Kodak employee named Earl Kage. He doesn't seem to appear in any of my paper literature. There is some online info, but aside from obits, etc., most of it seems to be related to things he (allegedly) has said to Lorna Roth, who seems to be a primary promoter of the "racist film" idea. Here is a quote from another website:

How Photography Was Optimized for White Skin Color

Earl Kage, former manager of Kodak Research Studios, told Roth that he didn’t remember any pressures from the Black community to improve their films.

The quote doesn't seem to indicate any problem, does it? The NYT article seems to paint a different story, though.

Anyway, I'm mainly trying to convince readers of the article to have a healthy skepticism toward the picture being painted in the article, not to mention the qualifications of the author in making such judgments.

Moderator Note: The factual,technical information that several experienced members submitted is quite clear. Info the author of the piece did not have. Writers produce products, for pay, to advance their careers, and /or to support their particular world view - popular views sell best.

This thread is not particularly attractive, photography oriented or useful for the site IMO. I try not to delete threads. Please just drop it.

'Shirley' type 35mm filmed charts were used to calibrate R, G, B values on professional (now mainly analog/legacy) telecines towards the end of the 1900s.

Kodak produced these to facilitate daily telecine line up and setting a standard base for consequent colour grading sessions.

When filming non caucasian actors it was the skill and craft of the DOP's to light them correctly due to the limited dynamic range of early colour emulsions.

21st century digital production cameras with their sophisticated LUTs (custom Look Up Tables) allow contemporary DOP's a range of colour palettes that handle most exposure situations.

My 55 + years in the photofinishing business qualifies me to speak about test images. I started using these materials in 1954, an after-school job, I was 16 years old. Photofinishing was big business, most everyone had a camera and their exposed films were taken to a camera shop or a drugstore for processing. In due course, I was Technical Manager for Eckerd Drugs headquartered near Tampa Bay Florida. I designed and built 7 huge photofinishing plants, each sized to develop and print 20,000 rolls of film a day (sized for holiday business). The Kodak Shirley test kit was an important tool. It came as a kit, a properly exposed color negative and a properly exposed and process print. It was a portrait scene crammed with colorful objects plus a scale of gray and colored tones. We reprinted the negative and compared our results to the suppled print. We often spent hours reprinting and adjusting the printer’s filters and software. This was daily Quality Control routine.

I could write volumes on the care and feeding of the high-speed printers that was the mainstay of the photofinishing business. It only took one carless mistake and thousands, perhaps tens of thousands of customer rolls of color film would print with an unacceptable bias. Such occurrences caused delay, we were charged to develop and print and return the package, next day. Re-printing to correct a quality control error could cost thousands of dollars. We did our due diligence to avoid.

While we did look at Shirley’s skin tone, our main attention was directed to the color and grayscale patches on the print. These we measured with a reflection densitometer. How well I remember the end-point values. We would find-tune to achieve 0.80 red – 0.80 green – 0.80 blue.

What I am telling you is, our attention was only incidentally on the skin tone, and our main objective was to get the 18% gray target “right”.

After I left Eckerd Drug, I was Technical Information Manager for Noritsu America. One department I managed had the responsibility to design, expose, and make test prints. This was a package that was completive to the Kodak test films. Our market was the one-hour photo-labs, worldwide.

The point I want to make – if photofinishing printers were adjusted to properly print these test materials, the vast majority of films, regardless of subject, printed with optimum color and density. This was due to the evolving refinements made to the exposure measuring mechanism and the application of this data in the control of the color and intensity of the exposing light. All this built into the photofinishing printers, regardless of who the maker was.

My 55 + years in the photofinishing business qualifies me to speak about test images. I started using these materials in 1954, an after-school job, I was 16 years old. Photofinishing was big business, most everyone had a camera and their exposed films were taken to a camera shop or a drugstore for processing. In due course, I was Technical Manager for Eckerd Drugs headquartered near Tampa Bay Florida. I designed and built 7 huge photofinishing plants, each sized to develop and print 20,000 rolls of film a day (sized for holiday business). The Kodak Shirley test kit was an important tool. It came as a kit, a properly exposed color negative and a properly exposed and process print. It was a portrait scene crammed with colorful objects plus a scale of gray and colored tones. We reprinted the negative and compared our results to the suppled print. We often spent hours reprinting and adjusting the printer’s filters and software. This was daily Quality Control routine.

 

I could write volumes on the care and feeding of the high-speed printers that was the mainstay of the photofinishing business. It only took one carless mistake and thousands, perhaps tens of thousands of customer rolls of color film would print with an unacceptable bias. Such occurrences caused delay, we were charged to develop and print and return the package, next day. Re-printing to correct a quality control error could cost thousands of dollars. We did our due diligence to avoid.

 

While we did look at Shirley’s skin tone, our main attention was directed to the color and grayscale patches on the print. These we measured with a reflection densitometer. How well I remember the end-point values. We would find-tune to achieve 0.80 red – 0.80 green – 0.80 blue.

 

What I am telling you is, our attention was only incidentally on the skin tone, and our main objective was to get the 18% gray target “right”.

 

After I left Eckerd Drug, I was Technical Information Manager for Noritsu America. One department I managed had the responsibility to design, expose, and make test prints. This was a package that was completive to the Kodak test films. Our market was the one-hour photo-labs, worldwide.

 

The point I want to make – if photofinishing printers were adjusted to properly print these test materials, the vast majority of films, regardless of subject, printed with optimum color and density. This was due to the evolving refinements made to the exposure measuring mechanism and the application of this data in the control of the color and intensity of the exposing light. All this built into the photofinishing printers, regardless of who the maker was.

Alan, although my days of shooting film pale in comparison with my time with digital, I read your post with a great deal of interest. I'ver never been a technical expert in either area of photography, and I learned a lot from you. For whatever it's worth, it also helped me gain more perspective on the OP.

Shadow detail is always harder than highlight detail.

 

I noticed in the article mention of "fair skin". What is so "fair" about it, and does that mean that

other skin colors are "unfair"?.

fair (adj.)

Old English fæger "pleasing to the sight (of persons and body features, also of objects, places, etc.); beautiful, handsome, attractive," of weather, "bright, clear, pleasant; not rainy," also in late Old English "morally good," from Proto-Germanic *fagraz (source also of Old Saxon fagar, Old Norse fagr, Swedish fager, Old High German fagar "beautiful," Gothic fagrs"fit"), perhaps from PIE *pek- (1) "to make pretty" (source also of Lithuanian puošiu "I decorate").

The meaning in reference to weather preserves the oldest sense "suitable, agreeable" (opposed to foul (adj.)). Of the main modern senses of the word, that of "light of complexion or color of hair and eyes, not dusky or sallow" (of persons) is from c. 1200, faire, contrasted to browne and reflecting tastes in beauty. From early 13c. as "according with propriety; according with justice," hence "equitable, impartial, just, free from bias" (mid-14c.).

Of wind, "not excessive; favorable for a ship's passage," from late 14c. Of handwriting from 1690s. From c. 1300 as "promising good fortune, auspicious." Also from c. 1300 as "above average, considerable, sizable." From 1860 as "comparatively good."

The sporting senses (fair ball, fair catch, etc.) began to appear in 1856. Fair play is from 1590s but not originally in sports. Fair-haired in the figurative sense of "darling, favorite" is from 1909. First record of fair-weather friends is from 1736 (in a letter from Pope published that year, written in 1730). The fair sex "women" is from 1660s, from the "beautiful" sense (fair as a noun meaning "a woman" is from early 15c.). Fair game "legitimate target" is from 1776, from hunting.

The following is for anyone interested in perusing info about the older negatives sometimes referred to as "Shirleys." If one reads the NYT article originally referenced, then reviews the instructions here, they can form their own opinions on the validity of the article.

I had saved a few older Kodak "Standard negatives," circa 1970s, for posterity. The attached photo shows how they came, with catalog numbers visible. They do not come with a guide print of any sort.

Here are some excerpts from the instructions for the VPSII package:

The Ektacolor negative set has similar instructions.

Moving forward in time to 2004, from an online Kodak catalog (L9, I think, probably findable on a search): the entry for Portra setup negs:

This is an exposure series, the set of negatives goes from 2 stops under exposed to 2 stops over. Again, there is no guide print included. Although another entry in the catalog DID include a guide print.

A few further comments, but first one should understand why there might be disagreement between Alan and I. Alan comes from a background of what is known as an "amateur lab." This does not mean that the employees are less qualified, but rather that the primary customers are amateur photographers. I think it is pretty well understood that the quality demands are not as high as they are with professional shooters, so no more need be said. This sort of processing is not my area of expertise but I'm not entirely ignorant; the outfit where I worked once owned the largest chain of mini labs in the US, where I was occasionally "loaned out" to assist with process problems or the like.

My background is more in the photography/process/materials end of the business, in an outfit that functioned more like a pro lab - portrait studio work shot on professional portrait films. In this sort of setup, the color requirements were tighter than a machine with exposure-measuring functions could do - everything had to be hand-balanced, etc. The exposure series of printer setup negs were nearly essential to 1) set up printer slope correction (largely to correct for paper reciprocity failure as exposure times vary) and 2) to link the printer settings to a video analyzer. Without this, one could still make high quality prints, it's just that at least one extra color test print would usually be necessary, which really slows down a high-volume operation.

As a note, our lab operation went through something like 600 to 800 slope negative sets per year, if that gives any idea of the work volume we did. And yeah, given that the Kodak list price was about $30-40 thousand dollars, we made 'em ourselves. 3 or 4 days work for one of my techs if everything went right, including reading a test density on every neg.

Ps, the system didn't use the size of image files I uploaded; if you can't read 'em you can "view image" then edit the url, changing the "md" near the end to "orig;" the orig is a little larger.

fair (adj.)

Old English fæger "pleasing to the sight (of persons and body features, also of objects, places, etc.); beautiful, handsome, attractive," of weather, "bright, clear, pleasant; not rainy," also in late Old English "morally good," from Proto-Germanic *fagraz (source also of Old Saxon fagar, Old Norse fagr, Swedish fager, Old High German fagar "beautiful," Gothic fagrs"fit"), perhaps from PIE *pek- (1) "to make pretty" (source also of Lithuanian puošiu "I decorate").

 

The meaning in reference to weather preserves the oldest sense "suitable, agreeable" (opposed to foul (adj.)). Of the main modern senses of the word, that of "light of complexion or color of hair and eyes, not dusky or sallow" (of persons) is from c. 1200, faire, contrasted to browne and reflecting tastes in beauty. From early 13c. as "according with propriety; according with justice," hence "equitable, impartial, just, free from bias" (mid-14c.).

 

Of wind, "not excessive; favorable for a ship's passage," from late 14c. Of handwriting from 1690s. From c. 1300 as "promising good fortune, auspicious." Also from c. 1300 as "above average, considerable, sizable." From 1860 as "comparatively good."

 

The sporting senses (fair ball, fair catch, etc.) began to appear in 1856. Fair play is from 1590s but not originally in sports. Fair-haired in the figurative sense of "darling, favorite" is from 1909. First record of fair-weather friends is from 1736 (in a letter from Pope published that year, written in 1730). The fair sex "women" is from 1660s, from the "beautiful" sense (fair as a noun meaning "a woman" is from early 15c.). Fair game "legitimate target" is from 1776, from hunting.

Thanks for the etymological/lexicognical lesson.

The following is for anyone interested in perusing info about the older negatives sometimes referred to as "Shirleys." If one reads the NYT article originally referenced, then reviews the instructions here, they can form their own opinions on the validity of the article.

 

I had saved a few older Kodak "Standard negatives," circa 1970s, for posterity. The attached photo shows how they came, with catalog numbers visible. They do not come with a guide print of any sort.

 

Here are some excerpts from the instructions for the VPSII package:

 

The Ektacolor negative set has similar instructions.

 

Moving forward in time to 2004, from an online Kodak catalog (L9, I think, probably findable on a search): the entry for Portra setup negs:

 

 

This is an exposure series, the set of negatives goes from 2 stops under exposed to 2 stops over. Again, there is no guide print included. Although another entry in the catalog DID include a guide print.

 

A few further comments, but first one should understand why there might be disagreement between Alan and I. Alan comes from a background of what is known as an "amateur lab." This does not mean that the employees are less qualified, but rather that the primary customers are amateur photographers. I think it is pretty well understood that the quality demands are not as high as they are with professional shooters, so no more need be said. This sort of processing is not my area of expertise but I'm not entirely ignorant; the outfit where I worked once owned the largest chain of mini labs in the US, where I was occasionally "loaned out" to assist with process problems or the like.

 

My background is more in the photography/process/materials end of the business, in an outfit that functioned more like a pro lab - portrait studio work shot on professional portrait films. In this sort of setup, the color requirements were tighter than a machine with exposure-measuring functions could do - everything had to be hand-balanced, etc. The exposure series of printer setup negs were nearly essential to 1) set up printer slope correction (largely to correct for paper reciprocity failure as exposure times vary) and 2) to link the printer settings to a video analyzer. Without this, one could still make high quality prints, it's just that at least one extra color test print would usually be necessary, which really slows down a high-volume operation.

 

As a note, our lab operation went through something like 600 to 800 slope negative sets per year, if that gives any idea of the work volume we did. And yeah, given that the Kodak list price was about $30-40 thousand dollars, we made 'em ourselves. 3 or 4 days work for one of my techs if everything went right, including reading a test density on every neg.

 

Ps, the system didn't use the size of image files I uploaded; if you can't read 'em you can "view image" then edit the url, changing the "md" near the end to "orig;" the orig is a little larger.

The following is for anyone interested in perusing info about the older negatives sometimes referred to as "Shirleys." If one reads the NYT article originally referenced, then reviews the instructions here, they can form their own opinions on the validity of the article.

 

I had saved a few older Kodak "Standard negatives," circa 1970s, for posterity. The attached photo shows how they came, with catalog numbers visible. They do not come with a guide print of any sort.

 

Here are some excerpts from the instructions for the VPSII package:

 

The Ektacolor negative set has similar instructions.

 

Moving forward in time to 2004, from an online Kodak catalog (L9, I think, probably findable on a search): the entry for Portra setup negs:

 

 

This is an exposure series, the set of negatives goes from 2 stops under exposed to 2 stops over. Again, there is no guide print included. Although another entry in the catalog DID include a guide print.

 

A few further comments, but first one should understand why there might be disagreement between Alan and I. Alan comes from a background of what is known as an "amateur lab." This does not mean that the employees are less qualified, but rather that the primary customers are amateur photographers. I think it is pretty well understood that the quality demands are not as high as they are with professional shooters, so no more need be said. This sort of processing is not my area of expertise but I'm not entirely ignorant; the outfit where I worked once owned the largest chain of mini labs in the US, where I was occasionally "loaned out" to assist with process problems or the like.

 

My background is more in the photography/process/materials end of the business, in an outfit that functioned more like a pro lab - portrait studio work shot on professional portrait films. In this sort of setup, the color requirements were tighter than a machine with exposure-measuring functions could do - everything had to be hand-balanced, etc. The exposure series of printer setup negs were nearly essential to 1) set up printer slope correction (largely to correct for paper reciprocity failure as exposure times vary) and 2) to link the printer settings to a video analyzer. Without this, one could still make high quality prints, it's just that at least one extra color test print would usually be necessary, which really slows down a high-volume operation.

 

As a note, our lab operation went through something like 600 to 800 slope negative sets per year, if that gives any idea of the work volume we did. And yeah, given that the Kodak list price was about $30-40 thousand dollars, we made 'em ourselves. 3 or 4 days work for one of my techs if everything went right, including reading a test density on every neg.

 

Ps, the system didn't use the size of image files I uploaded; if you can't read 'em you can "view image" then edit the url, changing the "md" near the end to "orig;" the orig is a little larger.