Revealing the Rainbow: An Introduction to the Science of Color

Susan Smith (National Postal Museum, Smithsonian Institution): Welcome everyone to today's talk, "Revealing the Rainbow: An Introduction to the Science of Color." I am Dr. Susan Smith, the Winton M. Blount Research Chair at the Smithsonian's National Postal Museum. I'm a historian, but I have been working recently with Smithsonian colleagues, Scott Devine, Ed Vicenzi, and Thomas Lam on a project studying the fading of stamps about which Thomas will present next week as part of the Fifth International Symposium on Analytical Methods in Philately, an event for which you can register online at analyticalphilately.org.

Now the four of us were looking specifically at the 1918 Curtis Jenny and its colors, red and blue. And on this team, I'm the furthest removed from science, and I was struck repeatedly during this project at how little I understood color and, particularly, how best to study it.

Happily, today's speaker will help illuminate exactly how to do that and introduce us to the science of color.

Olivia Kuzio is joining us today. She's a fellow in the Science Department at the Getty Conservation Institute in Los Angeles. As a color scientist working in a museum setting, the ways that she considers color in her work most often involve the material analysis and imaging of colorants, like pigments and dyes, that are used by artists. Her projects are centered around imaging systems, concentrating on the use and integration of reflectance imaging spectroscopies, macro X-ray fluorescent spectroscopies, and technical photography to generate distribution maps of the chemical, molecular, and structural components that comprise works of art.

In contrast, her recent dissertation research focused on developing and refining accessible strategies for carrying out advanced scientific imaging with more familiar, more affordable photography equipment. She enriched her graduate studies at the Smithsonian Museum Conservation Institute and the Getty Conservation Institute, where she did internships, and where she performs scientific imaging and material analysis on collections of objects ranging from ancient Near Eastern cuneiform tablets to Renaissance oil paintings to twentieth-century Bauhaus gouache color studies. She holds a Ph. D. in color science from Rochester Institute of Technology, and master's and bachelor's degrees in chemistry from RIT and Pennsylvania State University, respectively.

Olivia will be taking questions after her talk. You can type your questions into the Q&A, which you will find at the bottom of your screen. We will get to as many of them as possible before 5 pm. As this is a webinar, your cameras and microphones have been disabled and we will not be using the raise hand function. If you have any technical problems, please let us know through the Q&A, but unless there's a complete loss of sound, there will be very little we can do. You can choose closed captions by selecting show captions or the CC icon at the bottom of your screen.

Thank you again for coming.

Olivia.

Olivia Kuzio (Getty Conservation Institute): Wonderful! Thank you so much for passing the floor to me, Susan, and for that really kind introduction. I am thrilled to be here with you all today to reveal the rainbow and what will be a lightning speed about half-hour long crash-course introduction to the field of color science.

Before going full speed ahead, I'll share the framework of topics that I built this talk around to give you an idea of where we're headed. In a series of five sections, I'll discuss creating color, including each of the components that affect this process, their interconnections, and how together these result in the phenomenon of color. Describing color, using the intuitive vocabularies that help us make sense of what we see, and how these descriptions inform the organization of colors into scales and systems. Specifying color according to visual and numerical systems and the differences between these specification methods. Comparing color and using numerical methods to describe color differences in ways that are visually, visually meaningful. And finally, reflecting on the development of our understanding of color and considering how this understanding is continuing to evolve.

Seeing in color is a common human experience. Because most of us are so used to color being an inherent feature of the world that we navigate every day, I think it's easy to assume that it's a simple sensation.  It's probably most intuitive to describe color as belonging to objects. And, while it's true that coloration is due in part to the properties of colorants, which are the purely physical things that influence the color of materials, color is actually much more complicated than that simple assignment.

If you remember nothing else from today, remember this diagram: it shows that a color is a perceptual phenomenon that arises from a series of interconnected interactions that involve the modification of light by colorants and its subsequent detection and interpretation by the visual system.

Color is what we see, and it all culminates in the mind of the viewer. So, it's actually not that far off to say about color that it's all in your head. Understanding this, we can begin to more completely describe color by talking about each player in this set of interactions and we'll first start by considering the visible spectrum.

Visible radiation is just one form of energy within the electromagnetic spectrum that also includes radio, X-ray, ultraviolet and infrared radiation, but it's the only form of radiation to which the human eye is visually sensitive. It's a narrow band of wavelengths between about 400 and 700 nanometers nestled between ultraviolet and infrared. Another word for visible radiation is simply light, which for our purposes is best described by its wavelength in nanometers. One nanometer is one one billionth of a meter. So, we're talking about really tiny waves today. The hue that we recognize as blue is associated with wavelengths below about 480 nanometers; green, roughly between 480 and 560 nanometers; yellow, between 560 and 590; orange, 590 and 630; and finally red at wavelengths longer than 630 nanometers.

But now think about this: where is magenta? Magenta is an example of a common hue that's not an independent spectral color, meaning that there's no single wavelength of light that I could show you that would evoke the perception of magenta in your visual system. Instead, magenta-appearing light could be produced by mixing red and blue wavelengths from the extremes of the spectrum, and, in fact, most colored stimuli result from different combinations of many wavelengths.

And now that we've covered the basics of the visible spectrum, I can move onto illustrating that in graphical form, which looks something like this as we move into considering the role that light sources play in the creation of color. So, this plot here is called a 'spectral power distribution.' It shows the relative amount of each wavelength across the visible spectrum in,

in this case, daylight. The sun is a typical example of what we call a 'white light source.' These are the kinds of lights that run dispersed through a prism, like Isaac Newton did in the early eighteenth century, they separate out into the spectrum, and they provide a really beautiful example, beautiful visual evidence, that white light is actually a mixture of all of these wavelengths.

Another common example of a white light source is an incandescent bulb. While these kinds of bulbs also emit radiation across all of the visible wavelengths, the relative amount of the wavelengths in the incandescent spectrum are different than those in the daylight spectrum.

A typical incandescent lamp will look warmer because it contains relatively more of the redder wavelengths that appear compared with daylight, which is balanced more towards those bluer life wavelengths that appear cooler.

Last, we can consider the spectral power distribution of a red LED. This plot shows the characteristics of what is considered a narrow-band light source, because it emits radiation only within a small range, peaking at a single wavelength. So, in contrast with the 2 white light sources which emit at least some amount of radiation across all the visible wavelengths, because the red LED is completely lacking any emission in those shorter wavelengths, it definitely won't balance out to look white and, instead, appears intensely red.

Moving on, we can now consider the role that the materials with which light interacts play in the creation of color. There are a lot of things that can happen when light strikes an object, including absorption, transmission, and different kinds of scattering and reflection. And how these occur depends on things like the inherent physical properties of the material; variables, like the angle at which light strikes the material; and the roughness of its surface. That light - material interaction can get really complicated, really fast. And all of these events and factors totally impact the perceived color. But for simplicity, I'm only going to show here the effect that an object has on light in terms of what we call its 'spectral reflectance.'

Spectral reflectance plots describe materials in the same way that those spectral power distribution plots on the last slide, describing the light sources, did. In this case, these graphs show the relative amount of light reflected by the material at each visible wavelength. These are the color-coded spectral reflectance curves of several different paints that we can use to compare their reflectance characteristics and how these play a part in determining their perceived color. For example, the red, orange, and yellow paints strongly reflect longer wavelengths of light, but reflect very little of the shorter wavelengths. The difference between these three paint colors is the position in the spectrum at which this transition transition from reflecting very little to reflecting a lot of the incident light occurs. It's at a shorter wavelength for the yellow paint, longer for the orange, and longest for the red. Conversely, the blue paint most strongly reflects shorter wavelengths, and very little of the longer ones. While the magenta paint most strongly reflects both blue and red wavelengths, and at this point, as you've probably guessed, the green paint is strongly reflecting right in the middle of the spectrum.

And, just as a small aside for a little bit of context as to why I chose paints to illustrate the spectral reflectance of materials, I'm a color scientist who works in a museum and will often use the spectral signature or the shape of the reflectance curves of colorants in media, like paint, as a first means of identifying what a colorant is because each colorant has a unique spectral shape.

The final aspect to consider in the creation of color is detection by the eye and subsequent interpretation by the visual system. That latter interpretation part involving the upper parts of the visual system and the brain is ultimately responsible for the perception of color, but that system is hugely complicated. I totally encourage you to learn more about it, if you're interested in it, because it's really fascinating stuff, but I think that discussing it in depth is better left to experts in fields like physiology and neuro, cognitive, and vision sciences. What I will introduce here today is the trichromatic sensitivity of the color receptors in our eyes.

Our sensations of color are a result of having three types of these color receptors called 'cone cells,' which each respond differently to visible wavelengths. The letters L, M, and S are used to represent the three cone types according to their peak sensitivities in the long, middle, and short wavelength regions, respectively. I plotted what these spectral sensitivities look like on this graph, and, again, this is a similar graph to those on the last two slides, where it's showing the relative sensitivity of each cone type with respect to the wavelengths of the visible spectrum.

Note that the sensitivities of the cones are not evenly spaced across the visible range, and that actually the L And M cones overlap a lot with each other, but only a little bit with the S cones.

This uneven sampling of the visible wavelengths is pretty important, in particular, because it has some higher-level physiological implications on our ability to discriminate between colors.

The detection of light by the cones is a process of integration, which means that when light, having whatever particular spectrum it does, falls on them, the cones of each type sum up all the visible wavelengths to which they're sensitive. This reduces the spectrum of information into three signals, one from each cone type, and this is the reason that normal human color vision is called trichromatic. When viewing two colored stimuli like a pair of colored lights or colored objects, they'll match in color when they produce the same cone signals. This concept is called metamerism, and it's super important because it means that color matches can be calculated based on knowing the cone sensitivities across the spectrum and knowing the spectra of those stimuli. And, actually, this is the basis by which we can match colors across different media that have very different spectral characteristics. So, think of doing things like color proofing between physical prints versus digital displays.

Now that we've considered the basic processes by which color is created, we can sort of pivot to discussing the ways in which we describe and categorize color according to how it looks. We can first define a few one-dimensional psychological attributes that are useful to use when talking about color. These are hue, lightness, and chromatic intensity.

So, 'hue' is pretty intuitive; it's the attribute of a visual perception by which something looks red, yellow, green, blue, or a combination of adjacent pairs of these colors when you organize them in a closed loop. And that's the important thing about hue - it doesn't really have a start or an endpoint, so it just kind of makes the most sense to organize it as a circle.

'Lightness,' or what artists would call 'value,' is a color similarity to a neutral color ranging from black to white. So, the magenta color here is shown with respect to a neutral gray scale, where both are increasing in lightness from left to right.

And, finally, 'chromatic intensity' is simply defined as the amount of color. So, the salmon color is shown mixed with 2 different grays here and increasing in chromatic intensity from left to right.

When we arrange these scales in relation to one another as axes and organized colors along two or more of them at the same time, color order systems are born.

I'm gonna introduce 2 such systems, the first of those being the Natural Color System. The NCS has its basis in the ideas published by Ewald Hering in 1878, in which he refers to six elementary colors: white, black, red, yellow, green, and blue, and he calls these the natural colors.

Through the early and middle parts of the twentieth century, there were others who came along and interpreted and revised and produced physical examples of a system based on his ideas,

and these all eventually culminated in the Swedish Color Center Foundation releasing the Swedish Standard Color Atlas in 1979, which is the NCS physically realized. The central principle of NCS is defining colors by their similarity to Hering's natural colors. It's arranged in a double cone shape with NCS whiteness at the top, NCS blackness at the bottom, and NCS chromaticness radiating outward.

There's a key feature to NCS and that's that each hue has the same maximum chromaticness, which means that around the entire hue circle the highest defined chromatic strength of all hues is at the same level. We can visualize that hue circle by slicing through the 3D space horizontally. Hue is in turn defined as percentages of adjacent natural colors. For example, the hue YADR falls between full red at a hundred R and full yellow at a hundred Y in a proportion of 80 red to 20 yellow.

Slicing through that 3D space vertically lets us visualize what are called 'planes of constant NCS nuance.' Within this, a color's position is defined by percentages of its blackness and chromaticness like this square, which is at position 30% blackness and 50% chromaticness within that Y 80 R plane of nuance.

The NCS is produced in swatchbooks and fan decks at different glossiness levels, and it's often used to specify things like house paints and artist materials given that it's got that Swedish design-y type of flavor.

The second system I'll describe as the Munsell color system which Albert Munsell devised in 1905 as a teaching aid for art students. It was first published as the Atlas of Munsell Colors in 1915 and developed with emphasis on equal visual spacing and using instrumental measurements to define and create physical examples of the system. There are five further subdivided principal hues which he determined visually, rationalizing that these five should blend into a neutral gray when observed spinning on a disc.

Chromatic intensity is sampled along scales of constant value, which Munsell coined 'chroma scales.' The 3D representation of Munsell hue, value and chroma doesn't really have a defined shape like a cone or a sphere, because, unlike in NCS, the chromatic strength - that that chroma - isn't normalized between the hues. So, for example, a yellow at maximum chroma is much lighter than a blue at maximum chroma. So, the shape of the system is like kind of unwieldy and it's simply called the 'Munsell Color Solid.' The notation in the Munsell system is given as hue, then value over chroma, like so.

And it should just be noted that this system, too, has been continually refined and updated over the course of the twentieth century by other scientists and groups, like the Optical Society of America, and that it's also physically available in atlases of various versions. But perhaps, most importantly, this is the legacy for which my beloved Alma Mater, the Munsell Color Science Lab at RIT, is named.

Now we're going to move on to the specification of color. And, first, how this is done visually by using color mixing systems. So, these are systems that exemplify relationships between color primaries and mixtures of those primaries defining a color by the amount of each primary that's needed to meet it. For example, in run-of-the-mill color printing, colors are most often produced by mixing the four primary inks - cyan, magenta, yellow, and black. Color displays mix together three lights - red, green, and blue, and, in a more involved example, the Pantone Matching System is a much more complicated one that uses one or more of a set of fourteen inks to produce a single color.

The color recipes defined using these systems and others like them are largely responsible for making our manufactured world a vibrant one, but there are a few critical considerations to keep in mind when using these systems for visual specification. It's actually really difficult to highly, consistently produce visual examples of a specification across things like different inks, papers, and printing processes. And if this isn't done according to really controlled standardized procedures, big visual differences can occur in the physical products that come out of them.

Problems also arise when using specifications intended for one material to produce a different type of material. For example, using a printing specification to define a dye recipe. And, even if a match across materials is achieved, problems related to metamerism can arise if observation and illumination conditions aren't also specified and tightly controlled during that visual matching process. Because remember, for spectrally different samples, a meta, a metameric match is only guaranteed under identical viewing environments.

In short, physical examples of color mixing systems are guides - not truths - and using them properly for visual color specification requires a really high level of attention to those details at all stages of production and use. For exactly these reasons, there are a lot of applications for which visual color matching simply doesn't cut it. And so, over the past century or so, a threefold numerical solution called 'colorimetry' emerged to overcome limitations related to inconsistent measurement, illumination, and observation of colors during specification.

The first facet included developments and standardization related to the instrumental measurement of object spectral reflectance properties. The measurement of color is a topic worthy of several lectures in and of itself, and I think that recommending best practices for performing color measurement is best left to colormetrology experts who who really specialize in that field. But I'll say all the obvious things about, you know, good scientific data collection habits like representative sampling and repeating measurements, those all apply here.

The second facet included the standardization of lighting, where a number of light sources have been defined by the International Commission on Illumination, or CIE, for use in describing color. These distributions are formerly called 'CIE Illuminance' and are based on physical standards and statistical representations of measured light. There are standard illuminants for incandescent, fluorescent and daylight lighting, just to name a few, and some of the most commonly used in industry include D65, which represents indirect outdoor daylight, and Illuminant A, which represents incandescent light.

The third facet was the CIE's standardization of two observers, called the 1931 and 1964 standard observers. These observers are not real people, nor do they represent the visual sensitivity of any one person. Rather, they're the results of a series of complex visual experiments that were carried out at different points in the twentieth century and, for our purposes today, they can be thought of as mathematical models that are used to introduce consistency into the way that the numbers associated with perceived color signals are transformed into a numerical color specification system.

With those three important components in the creation of color standardized, there arose this novel, repeatable numerical means of specifying color called the 'CIE tristimulus system.'

With this colorometric methodology, we have yet another vocabulary for defining the colors of materials in this case by their tristimulus values, which can just be thought of as a three number coordinate in another one of these 3D spaces into which color can be projected.

However, in tristimulus space, colors are not equally visually spaced. This means that two colors that are numerically the same distance away from a third color, as defined by their CIE XYZ tristimulus values, might not appear to be visually the same amount different from that third color.

Whoops. So, there were efforts that were formalized by the CIE in the early twentieth century that So, they've been trying to develop new color spaces that overcome this problem and, therefore, better correlate with color perception.

The most well-known and widely used of these, what we call 'perceptually uniform color spaces,' is arguably CIELAB. CIELAB was officially born of the CIE's Color Imagery Committee in 1973, but is based on work by many scientists, including Adams, Nickerson, Glasser, and some others, some of which dates back to the early 40s.

In this 3D space, color is again defined by three values that are organized along a lightness axis, L*, and two perpendicular chromatic intensity axes, A* and B*. So, as not to error on equation and number overload just yet, I'll put it in words that the work that was done by the CIE Committee and all those scientists was to literally formulate the math by which colors in that perceptually non-uniform XYZ space can be translated into the coordinates of this more perceptually uniform CIELAB space.

Again, this is really important. It means that in this space, the positions of colors actually mean something visually. Colors that look to be the same amount different are also assigned by their coordinates to be about the same distance apart from each other within the space.

CIELAB isn't perfect, but it's still useful, especially with respect to small color differences, and it's still widely used in industry. Heck, even your Photoshop Picker will pick in lab coordinates, if you want it to. Nevertheless, perfect is the enemy of good enough, and perfect won the day in this particular battle when, toward the end of the twentieth century, the focus of the color community switched from developing color spaces with improved visual spacing to developing new ways to calculate distance within existing color spaces.

Until the nineties, the CIE's approach had focused on doing whatever math was necessary, no matter how complicated, to transform numerical color specification in XYZ to a more uniform color space, like CIELAB, in order to preserve the outcome that color differences within that transformed space could be calculated as the simple Euclidean distance between them.

So, this approach is realized in CIELAB as the Delta EAB equation, which defines the "as the crow flies" Pythagorean distance between two L*A*B* color coordinates. However, I've stated the visual spacing in CIELAB is not perfect. So, calculating numerical differences in this way, and assuming that they correlate with visual differences, no matter how far apart colors are or what region they're in, falls apart pretty quickly based on known perceptual irregularities in the space. As it turns out though, coming up with a better space than CIELAB and convincing people to drop CIELAB in favor of that new space are both really difficult tasks.

So, the solution became more math. I give you the CIE Delta E 2000 [CIEDE2000] formula which is terrifying, I know. I couldn't even muster the strength to type out the equations myself and the equation editor, so this is just a screen grab. So, we'll just call it what it is; it's a weighted color difference formula that is fit to the irregulars of the visual spacing in CIELAB. We'll describe what it does, which is spit out a convenient Delta E value that describes the difference between two colors in CIELAB, and we'll state why that's important, which is because it represents a further improvement of our ability to talk about color differences in a way that's visually meaningful and consistent.

Oh. Okay, we are nearing the end, friends. I have tried to weave throughout this presentation the names and accomplishments of some of the historic major players and overseeing bodies in the game of color and my explicit mentions here only went as far back as Newton and his prism in 1730, and they really only included a few of the many scientists, artists, and curious minds who have investigated color since. I need to acknowledge that there were so many others before and among these few that I've mentioned, who have also contributed greatly to the investigations and discoveries that have shaped our current understanding of color.

In case you didn't notice, I am also going to point out that I used the words twentieth century a whole lot, particularly particularly while speaking through the back half of my slide deck and covering the topics of organization, specification and comparison. This just goes to show that a framework for understanding the basics of what we consider the field of color science has only been developed and formalized in the past 100 years or so.

Color science as a discipline is relatively fresh and far from stagnant. It's fast growing, it's continually evolving, and there's tons of active research in domains that affect many different aspects of our daily lives. These include things like coatings and materials, displays, cameras and other imaging systems, augmented and virtual reality, and even things like social interactions, to name a few.

Color is anything but a solved problem in any of these domains and beyond. So, I really encourage all of you to leave here curious and to look a little more closely at all of the ways in which color makes your own life a little more vibrant.

With that I'd like to wrap up by extending my sincerest gratitude to Susan and Thomas for inviting me to speak today to the National Postal Museum for hosting this event, and to all of you out there, I can't see you, but I'm sure you've been a wonderful audience. I'm grateful that you simply showed up. I'm grateful for your interest and attention, and grateful for all of your forthcoming questions, which I'm sure will be thoughtful, but hopefully not too tricky.

Thank you.

Susan Smith: Thank you so much, Olivia! We do have quite a few questions.

Olivia Kuzio: Oh, my gosh!

Susan Smith: So first, where on the spectrum do black and white fall?

Olivia Kuzio: They do not. That's the simple answer and I guess there are different ways to

answer that question. If we're thinking about materials, black materials absorb all of the light that falls upon them. They don't reflect anything back to our eyes; therefore, we don't perceive anything, right, and white white materials reflect all of the light that falls upon them. Therefore, like we talked about, when all of those wavelengths mix together, we get the perception of white. So, that's one way to answer that question.

Susan Smith: Okay. You showed a slide with, well, on trichromatic sensitivity; does that slide help us understand color blindness, in terms of a lack of separation between the red and the green responses?

Olivia Kuzio: Sure. So, color blindness, true color blindness. Folks who are dichromats, we'll call them, are called dichromats because they are either missing or have irregularly formed

cone cells of one type. So, the most common types of color blindness come from L cones completely missing or not working correctly, or M cones completely missing or not working correctly. Some folks are missing their S cones, and they're another type of dichromat.

So, so yes, it's simply a question of how many kinds of signals you're comparing in your cone cells, whether that's three, a trichromatic system, or two, the dichromatic system.

Susan Smith: Okay, thank you.

How can a color be patented if they're all derived from known wavelengths?

Olivia Kuzio: Wow! That is, that's a doozy, isn't it?

I am no legal expert, so I don't know if I know how to answer that very well, but I'll think on it, and I'll ask people who might. And if folks want to get in touch with me offline, maybe I can point you to someone who could answer that.

Susan Smith: Nope, nope. Fair enough. Yes.

And these systems that you've talked about, are these used worldwide?

Olivia Kuzio: So, which systems?

Susan Smith: So, like, so for example, CIELAB.

Olivia Kuzio: Totally.

Susan Smith: The research on on remaking and making the system more nuanced and is that happening in research around the world? Or is this the type of work being done in certain places?

Olivia Kuzio: Absolutely. There are active color researchers all over the globe. When I, when I go to color science conferences to meet folks from quite literally everywhere, who are working on these kinds of problems. And yes, CIELAB is pretty much ubiquitous. That's why you see it in things like Photoshop, that, yeah, that's that's a ubiquitous program, right? And that's again, I think, I spoke to a little bit of the problem of maybe coming out with a new system that is used as widely as CIELAB because CIELAB is just so ingrained in the way that we talk about color, even though we know that it's not perfect.

But one of my favorite questions to ask people is, how good is good enough, you know, for your application and what you understand about CIELAB and and what you need it to tell you,

is it good enough? And in a lot of cases, it's good enough, and we need to think about, yeah, that idea of perfect being the enemy of good enough and getting by.

Susan Smith: So, what would it take to make the changes and developments widely accepted today, given all the research that's going on?

Olivia Kuzio: Sure. So, there are these big governing bodies that put out color standards. So, the one that I mentioned is obviously the CIE, the Commission International Commission on Illumination. Those are the folks who dictate from the top what color standards are and what color science means at a very formal level. And it takes a long, long time for color standards to get moved through that body, for everyone who's got their eyes on it to agree on the next generation of color science and how we talk about color. It's just a long process. Like any kind of big organization, it's slow moving, lots of eyes on it. So yeah, it would come from the top down. Probably. So, until the CIE comes out and says okay, this is what we're doing, and until people really get on board with that, we are using the systems that we're using and that I talked about today.

Susan Smith: Okay, and how did you become interested in color?

Olivia Kuzio: That is a great question. For me, color was a means to an end. I discovered the field of cultural heritage science and conservation science as an undergrad studying chemistry,

and when I knew that I wanted to stay in cultural heritage science, I knew that I needed a graduate degree. So, I was researching different graduate programs that would keep me aligned with my interest in this field, and it just so happens that there's a color science program

at Rochester Institute of Technology that wasn't that far away from where I wanted to be in the world at that time in my life.

So, I studied color because it helped get me here to to the museums, and I do not regret it at all. This is an amazing field. I truly knew nothing about it before I applied and went to graduate school. I mean, I did a dissertation on imaging, and I'm not even kidding when I tell you that I had not even touched a DSLR camera before I went to graduate school. So, I'm very fortunate to have learned as much as I did from the amazing folks in that program who really are the top-tier experts in this field.

Susan Smith: And how are you applying the color sceince and what you now know about color to your everyday work?

Olivia Kuzio: Sure. In a few different ways. So, the the science that I do, I'm actually looking at color and beyond, I'm studying the ways that different kinds of radiation interact with materials  in art collections. So, I look at the ways that infrared radiation and visible light information and ultraviolet and X-ray radiation, how all of those interact with materials, and how, like I talked about, when we use detectors to pick up those rays after they've interacted with material, what that tells us about what a material is made of. So, my my ideas about color have broadened. But I also am super fortunate to be working really closely with the photographers here at the Getty, super talented group of folks - I think they're online today - and we are working together to look more closely at the ways in which they manage color in their imaging workflows, because you can imagine that when we are taking images of art and putting them out there to the world, we want to make sure we're putting out images that look like what the actual art looks like. We don't want to be making representations of what what's in our museums and and hanging in the galleries. And also, if we're really careful about calibrating color in the images that we're taking today, we can image that same object ten years from now, calibrate the imaging color in a color-managed sense in the same exact way, and look at how colors might have changed in that time. So, it's really useful as a change detection method.

Susan Smith: And in that scenario what would you have to be controlling for? I mean the light in the room that you're doing this work in. What else?

Olivia Kuzio: Completely. Yeah. So, it's thinking about those those three aspects of creating colors. So, we want to standardize the light sources, we want to as best we can standardize the kind of cameras the detectors that we're using that that sub-in for our eyes in that situation, and we use things, so I don't have any pictures of them in this particular slide deck, but color checkers. So, so, those cardboard squares with lots of colorful squares on them that you see, in, you know, movie sets and in lots of imaging settings, but you might not necessarily know what it's for. Those are calibration standards for us to take a picture of. We calibrate that image of the color checker so that we know what the colors in it are, and then we can apply that same calibration to an image set later to know that we have done the same calibration set to both images now and later, and they are, therefore, comparable.

Susan Smith: Does that make it difficult to work with the newest technologies? I would think that these types of things would change. For example, the way that they resolve

Olivia Kuzio: It can be, but the calibration process does take care of some of the differences between detector technologies. Yeah, we'll just call it that.

Susan Smith: Okay.

Okay, is the Munsell system considered more accurate than some others, because the hues do not have the same max chroma.

Olivia Kuzio: I don't know that it's a case of more accurate. I, it depends on your goals and what you are using the system for. If you are - That's that's kind of a tricky question and maybe, can we come back to that? Can I think better about how to answer that?

Susan Smith: Sure. Sure. Okay.

A lot of philatelists use the Ridgway system. Is that one that you've heard of or come across? It was developed by the Smithsonian and wondering if anybody still uses it, if anybody else uses it.

Olivia Kuzio : I am mildly embarrassed to admit that I don't know the system.

Susan Smith: Fair enough.

Olivia Kuzio: but now I have reading to do when I hop off of this call. Could you tell me a little bit about it?

Susan Smith: I can't; it is also the first time I've ever heard of it, to be quite honest. So, the answer to that is, no.

In regards to stamps and colors - you know, we have a lot of philatelists on with us today - what type of light or lamp would you recommend keeping all of these questions of color in mind for

an average collector to use in terms of describing the color, to best describe the color?

Olivia Kuzio: Sure, I think the most important thing is that you're always using the same lamp. But after that you probably want to look for a lamp that's, want to look for a lamp that is made with these ideas of standard illuminance in mind. So, something that's got a spectral distribution that's close to one of these standard kinds of light sources that I talked about. And when you're using a light source that is spectrally similar to a standard, you can be relatively certain that other folks who are using lamps that are spectrally similar to a standard that we're all kind of

playing on the same field, at least with respect to illuminating the stamps, the collections.

Susan Smith: Sounds like we should be adding more descriptions of the equipment that we're using for some of these studies.

Olivia Kuzio: Completely. That is really important metadata to keep track of.

Susan Smith: So, I am told by my colleague that the Ridgway system was originally designed actually for ornithology.

Olivia Kuzio: Ooh, birds. Yeah. Yeah. Okay. I know folks who study color and birds. So, I'm gonna have to ask.

Susan Smith: And somebody else has added that Ridgway was an early twentieth century ornithologist whose book contained over 1,000 color chips.

Olivia Kuzio: Oh, fantastic!

Susan Smith: So, early philatelists used it, I gather, and they used the Ridgway assignations of names for colors, for shades, and that's how it entered into the

Olivia Kuzio: oh, okay, well, I'm interested in how it made the jump from feathers to stamps. That's some history of science context. So, I am looking forward to learning about.

Susan Smith: Absolutely, so someone else would like to know, if there's an easy way to explain this difference, between a colorimeter and a spectrophotometer.

Olivia Kuzio: A colorimeter is probably only going to tell you three numbers. So, it's gonna do what our eyes or what a camera does, and describe a color in terms of its, for example, L*A*B* coordinates from the CIELAB system, where a spectrophotometer is going to give you that full spectrum of information. So, how light is reflected off of the material you're measuring at every wavelength of the visible spectrum. So, actually, this is a good slide to have up for that question, I just realized I'm still showing this.

Susan Smith: Fantastic. Can you talk a little bit about additive and subtractive colors?

Olivia Kuzio: Sure. Additive colors, these are the kinds of colors that we think of as mixing light. So, when we're mixing light, that's how your display work, that's - let's maybe go to a different slide. Yes, I did not talk about this very much. Yeah. Mixing light. So, again, when you are exciting the visual receptors in the eye, using wavelengths across the entire spectrum those will mix to a white perception, whereas if we've got no light, you're gonna have a black perception. So, that's that's additive. Right? That's adding a blue light to a red light to make a magenta light.

Whereas in subtractive color mixing, we're thinking about the materials and how they absorb and reflect light. So, if I've got a material that looks red, it's absorbing all of the wavelengths that do not look red. Right? It's absorbing all of those blue wavelengths. So, if it's absorbing those and they're not making their way back to my eye, I'm not, I'm not going to perceive them. So, that's called subtractive mixing. That's when materials are subtracting away wavelengths that we're just not going to perceive then.

Susan Smith: Okay? And so maybe this is a good time to ask another question, what then would be the difference between blue and indigo?

Olivia Kuzio: This is a little bit semantic.

Susan Smith: Okay.

Olivia Kuzio: It depends on how you're using these words. Indigo is a, oh, I don't want to get it wrong on recorded video, a dye? A name for a dye, blue? So, I, yes, I hesitate to answer questions like these, you know, what's the difference between purple and violet?

As long as we all give some context and know what we're talking about then I think it is fine to use whatever word, whatever word floats your boat, but you should give some context for what you are talking about.

Susan Smith: So different categories, terms used from different categories?

Olivia Kuzio: Sure.

Susan Smith: From color science.

Okay, and I'm not quite sure this question, can you discuss CQS and CRI as determining accuracy of light sources and how it relates to reflectance?

Olivia Kuzio: Ah, so these are some of the standards for color rendering, and I have to admit, I do not know a lot about those topics. But what I do know is that those kinds of metrics are ways to to put numbers to how well a light source will render a certain color. So, it has to do with

the interaction of the spectral content of the light source, and whether that makes a color look like what we would expect it to look like. But, again, it's not a topic I'm super-well versed, and so I kind of hesitate to say more words than that.

Susan Smith: Fair enough.

In the world of art conservation, do art conservators try to assign a color or shade name to a CIE or a CIELAB value that you've determined by instrumental methods?

Olivia Kuzio: In my, in my experience, no, we we don't really talk about colors like that, but to be fair, I don't really know the processes by which, like, the painting conservators are are mixing their paints. I think that it's done visually. They're quite talented and quite well trained in in that kind of visual matching.

In in my particular experience, in my conversations, I don't think so.

Susan Smith: Okay. Alright, I'm sorry there's so many questions here. I'm trying to vet them. So, okay, let's take a an easier one, I hope.

What's your favorite color?

Olivia Kuzio: Green, of course, because it’s the best. No, but truly green. For for a very long time, I drove a bright lime green Jeep Wrangler that I adored a lot because it was bright lime green.

Susan Smith: Sounds very special!

Okay, given your studies of art and your study of color, and I realize this is, you know, fairly recent in terms of your time at the Getty, do you have any quick observations or comments about how you're pulling together your education, what you're learning at the Getty, this study of color science?

And you had mentioned to me, and as I mentioned to everybody else that your dissertation work was quite different from what you're doing now in terms of what you had at your disposal to study with, right? So, you looked at more affordable means in the dissertation and then had the, you know, the opportunity to write about that and think about what somebody without access to these labs would be able to do.

Can you comment on that in relation to the study of color and your study of art?

Olivia Kuzio: Sure.

I'm so fortunate to work where I do. The Getty has the most amazing access to different tools, technologies, collaborations, when when we don't have particular tools and technologies here, which is rare. So fortunate. So, the things that I think about are, you know, yeah, we have access to all those things, but we have the same kinds of struggles as other institutions and there are processes that we are striving to do better at. And there are simple solutions in a lot of cases, whether it's just, you know, purchasing new color targets for all of the imaging studios to make sure that we have the newest standards, that they're not dusty and old and dirty, and throwing off our calibrations.

So, it's, I think, paying attention to small details and paying attention to small details like that applies whether you're at a big institution or you're a single photographer working at a small institution. A lot of times, it's process control and workflow development, and sticking to those processes and workflows, and working in a consistent manner that means more to to the quality of work and the ways that we're documenting, archiving, studying our collections, than having the latest and greatest technologies.

Susan Smith: Go ahead, Thomas.

Thomas Lam (Museum Conservation Institute, Smithsonian Institution): We've a question here that I don't want to have missed because I accidentally miss hit the button. So, want to make sure it gets answered. How does a color's gloss get measured in the L*A*B* system?

Olivia Kuzio: The short answer is that it doesn't. In the L*A*B* system, ideally, you would be considering the spectral reflectance of a material as measured in a diffuse sense, so excluding that specular, glossy component but there are ways in which gloss is measured. There are special instruments, glossmeters that we can use to measure the glossiness of materials and

material appearance and color.

Material appearances is a whole huge field that I did not even touch on today. Color is so complicated, people; this is really just brushing the surface, and these are sort of high level topics that I've covered. But yes, of course, things like the glossiness, the roughness of a surface affects the perceived color, but the measurement of color and the spaces in which we talk about that CIELAB is not an appropriate place to talk about color glossiness.

Susan Smith: What are the next big frontiers in color science?

Olivia Kuzio: Hmm! Well, we can - let me just switch a slide really quick here. This one.

So, so talking about, you know, glossiness and how materials and the materiality of things affects color. I really only covered today this interaction, right? So, light interacting with colorants, this arrow. And I talked about this arrow; so, after light bounces off an object and object has affected the light, how we perceive that, and what that means. I did not really touch on this arrow. This arrow would refer in this diagram to all the other stuff that's going on in your environment. I mean, there's a ton of light bouncing around in in this little conference room that I'm in that's also affecting my visual perception of the screen I'm looking at. When I'm looking between the screen and looking at the windows, that's palm trees outside. When I'm looking at this screen versus the screen that I'm projecting to, so that I can have notes here, and you all are looking at my slides there.

All of these kinds of things are, well, not all of them, but a lot of them are considered in these things called 'color appearance models.' And these are color spaces that enable color prediction based on knowing those extraneous viewing and illumination conditions that are also affecting our color perception in in a situation. So, color appearance modeling is like the the next stage, the newest and greatest way of defining and talking about color beyond just simple, three spaces like CIELAB, and thinking about how color appearance modeling is going to

make our ability to do things like AR and VR a lot better. So, you know, if you put on those Apple AR augmented reality glasses, what does it mean to be perceiving the outside world and the things that are physically around you, but also having things in color displayed on the glasses that are close to your eyes.

All of this is gonna, hopefully, be wrapped up in color appearance modeling that colleagues of mine at RIT in the Munsell Color Science Lab. They're thinking about problems like this. So, that's a really particularly interesting problem to me. It's something I don't know a lot about from my own work, but from the outside, looking at him. Wow! That is hugely complicated. And I'm glad that there are other people who are thinking really carefully about those kinds of problems.

Susan Smith: Wonderful. Thank you so much for doing this, Olivia, and for joining us and explaining color, and introducing us to the science and things to think about. I appreciate it!

Olivia Kuzio: You are so welcome. Thank you for having me.

Susan Smith: It's been wonderful.

Olivia Kuzio: This has been a ton of fun. I've been looking forward to it, and I'll keep thinking about this stuff, and maybe, just let me say, please do reach out. Look me up. Here's my name and my affiliation again for all of you online. I would really love to hear from you

Susan Smith: Thank you, Olivia, and thank you for all of you for coming. Take care!

Olivia Kuzio: Thanks, everyone.