“Color science”

[aizan]

What is the definition of everybody’s new favorite photo term, “color science”? Its origin seems to be in Canon marketing materials from a couple years ago, and refers to the “secret sauce” behind the “look” of their sensors.

 

[Chriscrawfordphoto]

Studying color is part of studying light, something physicists do.

 

[Dwig]

Studying color is part of studying light, something physicists do.

Actually, that is only half of the real science. There is an equally large, if not larger, field of study involving the perception of light. So we have physicists studying light and we have biologists and psychiatrists studying how animals, humans in particular, perceive light, from the receptors in the eye to the massive computer (read: brain) image processing that is also involved.

 

[jsrockit]

What is the definition of everybody’s new favorite photo term, “color science”? Its origin seems to be in Canon marketing materials from a couple years ago, and refers to the “secret sauce” behind the “look” of their sensors.

Yeah, so many people use this... in the last year, so much more. Just as annoying as all other photographic equipment buzz words.

 

[ptpdprinter]

Is is a buzzword to explain why different sensors in different cameras yield different colors in the same scene, i.e., why RGB in one differs from RGB in another. I am not sure how much of the difference results from conscious decision making, and how much results simply from different manufacturing technique. In a perfect world, on the digital side, there would be no variation.

 

[skucera]

Actually, that is only half of the real science. There is an equally large, if not larger, field of study involving the perception of light. So we have physicists studying light and we have biologists and psychiatrists studying how animals, humans in particular, perceive light, from the receptors in the eye to the massive computer (read: brain) image processing that is also involved.

So true! I was a "color science" test manager for a company that made spectrophotometers, colorimeters, and software for doing color matching in many industries. I was the odd man out because I had two art degrees, but we had four PhD's on staff: one with his PhD in color science with decades of experience at Kodak and RIT, one with a PhD in architectural engineering who later went to work at Pixar who was an excellent mathematician, one with a PhD in criminology who was hired because he was an excellent programmer, and the site manager with a PhD in electrical engineering with dozens of patents to his name. Perceptual psychology was the very core of our work, and we had fun discussions about the changes in color science over the decades as the renowned researchers in the field aged, and their own perceptions of color influenced their findings. I have so many good memories from that time.

Scott

 

[lynnb]

I understand it to mean the colour output from the combination of sensor, firmware and software - for example with Canon, how the Canon-designed sensor data is manipulated by firmware inside the camera (to create a RAW file and also JPG) and then to final output through Canon's own DPP software. This capture/processing imaging chain gives the "Canon look", particularly marketable for skin tones. Canon calls this their color science. Nikon has their own imaging chain processes to give the "Nikon look" to the files.

I think this is the generally understood meaning of the term "colour science".

Postscript: Goethe's Theory of Colours (first published 1810) was an interesting early exploration of colour harmony and aesthetics.

 

[skucera]

In my haste I forgot to answer Aizan's question. Sorry.

Color science is the study of human perception of color, but the practical application is in methods to match colors, so what you capture in your camera can be faithfully reproduced on your display and printed on paper. Back in 1997-2002 when I worked in color science, the main customers for color matching were on-line retailers who wanted to avoid returns because the color of clothing didn't match what users saw on their computer screens, graphic designers in many industries (print, TV, on-line), and medical imaging. "Color science" is definitely not an invention of Canon as a sales gimmick.

Perhaps it is no surprise, but Isaac Newton started studying color when he first used a prism to break white light into a rainbow, which he discovered could not be refracted further. Others studied color over the centuries, but a surprisingly close understanding of human color perception came from an American, Albert Munsell. You can read a little history about color science history at:
https://munsell.com/color-blog/history-of-color-systems/

Another milestone was the work of the French CIE organization:
https://medium.com/hipster-color-science/a-beginners-guide-to-colorimetry-401f1830b65a

Anyway, there is a lot of science to read if you're really curious. I started studying color science in college, and continued it on the job later.

Scott

 

[lynnb]

Thanks for the links Scott, it's a subject I find very interesting.

 

[willie_901]

Color science is very complicated. Since it's a science it's also a data-driven objective pursuit.

Subjective color perception is an entirely different subject. It is ad-hoc. However, this doesn't mean individual preferences aren't valid. They are. They are important as well.

 

[aizan]

I wasn’t talking about real color science, but rather the ineffable, proprietary, branded “color science,” e.g., “Fuji color science.” The assumption is that the color output of one camera brand DOESN’T match the color output of other brands, or that it is too much work to make them match. Nobody would ever say that actual color science was brandable.

One question could be “how does real color science relate to branded ‘color science?’”

 

[Rob-F]

Actually, that is only half of the real science. There is an equally large, if not larger, field of study involving the perception of light. So we have physicists studying light and we have biologists and psychiatrists studying how animals, humans in particular, perceive light, from the receptors in the eye to the massive computer (read: brain) image processing that is also involved.

Yes, we studied light and color when I was taking my clinical psychology courses--as well as in a color photography course I took. The instructor in the photo course stressed that "color isn't out there in the world; it's in your head." (but of course, differences among cameras are in the cameras, not in our heads.)

 

[aizan]

I understand it to mean the colour output from the combination of sensor, firmware and software - for example with Canon, how the Canon-designed sensor data is manipulated by firmware inside the camera (to create a RAW file and also JPG) and then to final output through Canon's own DPP software. This capture/processing imaging chain gives the "Canon look", particularly marketable for skin tones. Canon calls this their color science. Nikon has their own imaging chain processes to give the "Nikon look" to the files.

That's as good a definition as I've seen so far. Is there something in that chain of processes that makes the color science of one camera brand inimitable? Wasn't the purpose of ColorCheckers to make the color output from various cameras all look the same? 😀

 

[Ko.Fe.]

Among professionals I'm involved with it is called as color reproduction.

 

[skucera]

That's as good a definition as I've seen so far. Is there something in that chain of processes that makes the color science of one camera brand inimitable? Wasn't the purpose of ColorCheckers to make the color output from various cameras all look the same? 😀

Yes, the purpose of color checkers is to measure the difference between what your camera or scanner has captured, then to evaluate scientifically how it differs from the known color of the patches on the chart, so you can compensate to reproduce the exact colors that should have been captured. There is both science and art in doing that.


It isn't much of a secret that all these color matching systems still leave room for biases in the engineering, and room for both company color signatures and for cultural differences around the world. What do I mean by that? Traditionally, Kodak color had a warmness to photos recorded on Kodak films... more golden skin tones, and higher color saturation in mid-tones. The drawback was that Kodak color was optimized for white folks, and pretty fair complected white folks at that. Skin tones for people of darker complexions was too saturated, and folks with really dark complexions really didn't look natural at all. Color balancing in Europe was cooler in general (perhaps reflecting the warmth of sunlight their more northerly latitudes than the US, but that is debatable), and that could be really seen when comparing American photography magazines to British or European camera magazines on the newsstand. There's still a lot of social commentary on these color matching system biases, and some lingering anger. Let me know if you'd like some links on this subject, but is a touchy one for some folks, and I'd prefer that forum members search for these articles on their own. They're illuminating both about color science and about history too.


Scott

 

[willie_901]

One question could be “how does real color science relate to branded ‘color science?’”

I never heard the term color science as a marketing tool before this thread.

I have heard many photographers express strong personal preferences for their camera's JPEGs' perceived color aesthetics. Occasionally people will fell the same way about rendered raw files.

In terms of branding, it is possible some brands spend more R&D to optimize color rendering than others. For instance, we all are aware of the importance of the IR filter layer.

Another example is FUJIFILM's Electronic Materials business. One of this group's products are "primary color filters for image sensor applications". Specifically, "5th generation COLOR MOSAIC® is designed for high-end, high-resolution CMOS and CCD color sensors". So, this brand does invest in science and engineering to develop sensor color-filter arrays. Whether or this results in an actual advantage is another matter. Also, it seems reasonable to assume FUJIFILM sells CFA film (not necessarily manufactured array assemblies) to other brands, but who knows?

A brand's proprietary, in-camera demosaicking algorithms are also important.

 

[Roger Hicks]

It's like Pentax's "Super Multi Coating", a marketing adaptation of something real. Because of misleading publicity, lots of people believe that Pentax invented multi-coating. They didn't. Zeiss were experimenting with it during WW2 and Leitz was using it on production lenses in the 1950s.

Cheers,

R.

 

[ASA 32]

The deep knowledge and wide variety of interests of RFF members is truly wonderful. One of the reasons that, yes, RFF is indeed relevant.

 

[jbielikowski]

Yeah, so many people use this... in the last year, so much more. Just as annoying as all other photographic equipment buzz words.

try to watch any review on YT, in the latest fullframe blitzkreig it's the most annoying "feature", guess nobody cares for megapixels anymore so they "invented" this.