Daniel sent us this one — he's been thinking about color, specifically the methods we use to identify colors with precision. We all know RGB for screens, but what about when it really matters? Like, say, the exact shade of blue on a national flag. Is there a single authoritative color scale that everyone defers to when precision is non-negotiable? And if you need to measure that color in the real world, is a colorimeter the right tool? It's one of those questions that sounds simple until you actually try to answer it.
It sounds simple until you realize the blue of the European Union flag is Pantone two eighty six C. But the blue of the French flag? Almost fractally complicated.
Of course it is. The French would philosophize their own tricolor.
They're not alone. We assume color is objective — that "red" is "red" — but the moment you try to specify it across materials, across lighting conditions, across decades of fading and manufacturing variance, the whole thing fractures. And with global supply chains and brand consistency demands, the question of which color scale is actually authoritative has never been more practically urgent. When a government or a manufacturer or a designer needs to stipulate a color with zero ambiguity, what system do they reach for?
The answer, I'm guessing, is not "just one.
It's a patchwork. And to understand why, you have to contrast the digital world with the physical world. In digital, we've got RGB, hex codes, the sRGB gamut — it's all additive light, mixing red, green, and blue pixels. But the physical world is pigments and substrates and lighting conditions. You're not emitting light, you're reflecting it. And that translation problem — from a screen to a swatch to a painted steel beam to a nylon flag — that's the hidden complexity.
When you need to tell someone in a factory in Vietnam exactly what blue to use on a flag that's going to fly outside the UN headquarters, and it has to match the one that was made in Portugal three years ago, what do you reach for?
That's the question. And there are three major contenders. You've got Pantone, which is proprietary and design-centric. You've got RAL, which is the German industrial standard used heavily in architecture and powder coating. And you've got NCS, the Natural Color System, which is Scandinavian and based on human perception rather than ink mixing. Three completely different philosophies about what color even is.
Let's start with the one most designers know best — Pantone — and then see how it holds up when you leave the design studio.
The Pantone Matching System, PMS, is fundamentally a proprietary numbering system tied to physical swatch books. And this is the first thing most people get wrong. Pantone colors are not spectral definitions. They are not measurements of light. They are recipes. Specific ink mixtures on specific paper stocks. The swatch book you buy — and they're not cheap, we're talking hundreds of dollars and they recommend replacing them every year because the inks fade — that book is the standard. The number is just a reference to the physical swatch.
It's a cookbook, not a thermometer.
And that's both its strength and its weakness. The strength is that a printer anywhere in the world can look up Pantone one eight five C, find the recipe — thirteen parts Rubine Red, three parts Warm Red, whatever the mix is — and produce something that matches the swatch. The weakness is that this only works for offset printing on coated paper under controlled conditions. The moment you change the substrate — fabric, plastic, metal, uncoated paper — the same ink recipe produces a different perceived color.
Which is why Pantone has multiple books. Coated, uncoated, textiles.
Over three thousand colors in the current PMS library, but only about eleven hundred are achievable in textile under the Fashion Home and Interiors system. And here's where it gets really interesting from a business perspective. The twenty twenty-two Color of the Year, Very Peri —
The periwinkle that launched a thousand brand decks.
It was a color that didn't exist in the Pantone library before they announced it. They literally created a new ink formulation, gave it a name, declared it the Color of the Year, and suddenly every designer in the world needed to buy the updated swatch book to have access to it. It's brilliant ecosystem engineering. Create demand for your proprietary standard by making the standard a moving target.
The iPhone strategy, but for color.
It reveals something important. Pantone is a commercial product, not a scientific standard. It's optimized for the graphic design workflow. It dominates that world because it solved a real coordination problem between designers and printers. But it was never designed to be universal.
What about when a designer tries to use Pantone outside that workflow? Say you're an industrial designer and you spec a Pantone color for a plastic injection-molded part. What happens then?
You get a very frustrating phone call from your manufacturer. Because Pantone doesn't tell you how to achieve that color in polypropylene. The ink recipe is meaningless — you're not printing, you're mixing colorant into molten plastic. The manufacturer has to do a spectral match from scratch, essentially reverse-engineering what combination of pigments in that specific polymer will produce something that looks like the Pantone swatch under the agreed-upon lighting. And because plastic has a different surface gloss and subsurface scattering than coated paper, it might never match perfectly from every angle.
The designer thinks they've given a precise specification, but they've actually just handed over a vibe.
A very expensive vibe. And this is where you see companies develop internal standards that sit on top of Pantone. Apple, for instance, doesn't just say "space gray is Pantone Cool Gray eleven C." They have proprietary anodization specifications that achieve a specific spectral reflectance on aluminum. The Pantone reference is a starting point for conversation, not the final word.
Which brings us to the Germans.
Founded in nineteen twenty-seven by the German Institute for Quality Assurance. The Reichs-Ausschuss für Lieferbedingungen, if you want the full original name. And RAL takes a fundamentally different approach. Instead of ink recipes, RAL colors are defined by spectral reflectance curves. A RAL color specification is essentially a graph — at each wavelength of visible light, from about four hundred to seven hundred nanometers, here's what percentage of light this surface reflects.
It's substrate-independent. The same spectral curve can apply to paint, powder coating, plastic, fabric — you just need to formulate a material that hits that curve.
That's the critical technical distinction. RAL ninety ten, Pure White, is probably the single most specified color in European construction. You'll see it on window frames, radiators, architectural metalwork. And the specification isn't "mix this much titanium dioxide into this much binder." It's "the finished surface must reflect light according to this spectral curve." How you achieve that is your problem.
Which is much more flexible but also harder to guarantee. An ink recipe is an ink recipe. A spectral curve requires measurement.
And RAL has exactly two thousand five hundred thirty colors as of twenty twenty-six, each with a unique four-digit code and a documented spectral reflectance curve. The system is heavily used in architecture, industrial coatings, powder coating, and traffic signage across Europe.
There's a story I love about RAL that illustrates why this approach matters. The German Autobahn signs. Those iconic blue highway signs — they're RAL five thousand two, Ultramarine Blue. And the specification isn't just the color when new. It includes acceptable fade tolerances over a fifteen-year service life. The sign manufacturer has to demonstrate that after the equivalent of fifteen years of UV exposure, the spectral curve still falls within a defined envelope.
That's a perfect example of industrial thinking versus design thinking. Pantone says "here's the color." RAL says "here's the color, and here's how much it's allowed to drift over time, and here's how to measure whether it's still in spec." It's a maintenance standard, not just a production standard.
Then there's the Scandinavians, who looked at all this and said, "What if we based it on how humans actually see color?
The Natural Color System, NCS, developed by the Scandinavian Color Institute. It's built on the Hering opponent-process theory of color vision — the idea that our visual system processes color in opposing pairs: red versus green, yellow versus blue, black versus white. An NCS notation encodes four values: hue, blackness, chromaticness, and whiteness. So a color might be notated as something like S twenty thirty dash Y ninety R, which tells you it's a yellowish-red with a specific proportion of blackness and chromaticness.
It reads like a license plate, but the logic is elegant. You're describing the color as a human perceives it, not as an ink mixer produces it or a spectrophotometer measures it.
That's why NCS is huge in architecture and automotive interiors in Northern Europe. When an architect specifies a wall color, they're thinking about how the space will feel to a person standing in it, not about the reflectance curve of the paint. NCS maps more directly to that design intuition.
How does that work in practice, though? If I'm an architect and I walk into a space and think "this wall needs to feel less black, more chromatic," how does NCS help me communicate that?
The notation makes it explicit. If you have a color notated as S twenty thirty dash Y ninety R, and you want it to feel less black, you reduce the blackness number — maybe S ten thirty dash Y ninety R. The system lets you adjust perceptual dimensions independently in a way that Pantone numbers don't. A Pantone number is just an index. It doesn't encode any information about the color's relationship to other colors. NCS gives you a coordinate system for color perception. You can say "same hue, same chromaticness, but half the blackness" and the notation reflects that.
Three systems, three philosophies. Pantone says color is a recipe. RAL says color is a physical reflectance curve. NCS says color is a perception. And somehow, national flags have to navigate all of this.
This is where it gets wonderfully messy. Let's take the United Nations flag. The blue is specified as Pantone two seventy nine C. But the United States State Department specifies the American flag in both Pantone and CIE Lab coordinates — that's the device-independent color space defined by the International Commission on Illumination. They're hedging their bets.
Covering the covers.
The United Kingdom's Union Jack has no official Pantone specification at all. It's defined by spectral data in British Standard BS three eighty one C. The colors are named — Union Flag Blue, Union Flag Red — but the actual specification is a spectral reflectance curve, not a Pantone number.
The Canadian flag? I remember reading something about them having multiple official versions.
At least four. The Canadian flag has official color specifications for fabric, for paint, for digital display, and for embroidery. Different materials, different viewing conditions, different standards. And here's the thing — they don't all match exactly. They can't. The red that works on a nylon flag flapping in sunlight is not the same red that works on an embroidered patch on a uniform under fluorescent lights.
Which brings us to metamerism. The hidden gremlin.
Metamerism is the phenomenon where two colors match under one light source but not under another. You've experienced this — you buy a pair of socks that look navy blue in the store, and then you walk outside and suddenly they're purple. The spectral reflectance curves of the two dyes were different, but under the store's lighting, your eye couldn't tell them apart. Change the illuminant, and the mismatch reveals itself.
A flag that looks perfectly correct under D sixty five daylight — which is the standard illuminant representing noon sunlight — might shift noticeably under fluorescent office lighting or LED streetlights.
This is why national flags often have multiple official specifications for different viewing conditions. The flag code might say "under CIE illuminant D sixty five, the red shall have these Lab coordinates, and under illuminant A — which represents incandescent light — it shall have these slightly different coordinates." They're acknowledging that color isn't a property of the object. It's a relationship between light, surface, and observer.
I want to linger on metamerism for a second, because it's one of those concepts that sounds academic until it bites you. There's a famous cautionary tale from the automotive industry. A car manufacturer — I won't name names — had a bumper cover molded in one factory and the metal body panels painted in another. Both matched the same Pantone specification perfectly under the factory's inspection lighting. But under the sodium-vapor lights of a dealer lot at night, the bumpers looked like they were a completely different shade. Customers were bringing cars back thinking they'd been in unreported accidents.
That's the nightmare scenario. And the root cause is that the pigments used in the flexible plastic bumper cover had a completely different spectral reflectance curve than the pigments used in the rigid metal body paint. They were metameric matches — identical under one light, divergent under another. The only way to catch that before production is with a full spectrophotometer reading and a metamerism index calculation. A colorimeter alone won't see it coming.
We've got three competing systems, each with its own logic. But knowing which system to use is only half the battle. You also need the right tool to measure the result. Which gets to the second part of the prompt — is the colorimeter the tool of choice?
This is where I need to make a distinction that a lot of people miss. A colorimeter and a spectrophotometer are not the same thing. A colorimeter measures color using filters that approximate the response of the human eye. It captures what are called tristimulus values — X, Y, and Z — which represent how the three types of cone cells in your retina would respond to that light. It's fast, it's relatively cheap, and it gives you a number you can work with.
It's approximating.
It's approximating the human observer. And the standard observer functions that colorimeters are built around — the CIE nineteen thirty-one standard observer — were based on experiments with a two-degree field of view. That's roughly the size of your thumbnail at arm's length. It's a very narrow window of color perception.
Which got updated recently.
Twenty twenty-five. The CIE zero fifteen twenty twenty-five standard updated the observer functions for the first time since nineteen thirty-one, incorporating a ten-degree observer for peripheral color matching. Because it turns out your color perception changes depending on whether you're looking at a tiny swatch or a whole wall.
The colorimeter is giving you a measurement based on how a hypothetical human eye from nineteen thirty-one would see a two-degree spot of color. Under a specific light source. Change the light, change the measurement.
That's the key limitation. A colorimeter gives you the color under a specific illuminant. A spectrophotometer, on the other hand, measures the full spectral reflectance curve — at ten-nanometer or twenty-nanometer intervals across the visible spectrum. Instead of three numbers, you get thirty-one or more data points. That spectral curve is the gold standard for industrial color matching because it captures everything. You can mathematically predict what the color will look like under any illuminant. You can detect metamerism before it becomes a problem.
If you're specifying the paint for a Boeing sensor cover, you want a spectrophotometer.
That's not a hypothetical. There was a situation during the twenty nineteen Boeing seven thirty-seven MAX grounding where a mis-specified paint color on a sensor cover — I've seen references to RAL nine thousand five versus Pantone Black six C — contributed to a readability issue in low light. When you're dealing with safety-critical components, the difference between "black" and "black" can matter enormously if the spectral reflectance in the near-infrared is different and the sensor is looking in that range.
Wait, walk me through that. How can two blacks that look identical to a human be different to a sensor?
Because "black" to the human eye just means low reflectance across the visible spectrum, roughly four hundred to seven hundred nanometers. But a sensor might be operating in the near-infrared, say eight hundred fifty nanometers. One black pigment might absorb heavily in the near-IR while another reflects strongly. To your eye, they're both black. To the sensor, one is a mirror and the other is a void. If the sensor is an optical proximity detector and it's expecting a certain reflectance in the near-IR, the wrong black paint can render it effectively blind.
The colorist's skill, then, isn't just knowing which pigments to blend. It's knowing how to hit a spectral target without introducing metameric failure.
The practical workflow in a serious industrial setting goes like this. A manufacturer receives a specification — say, a Pantone number. They don't just mix the ink recipe. They mix it, apply it to the actual substrate, then measure the result with a spectrophotometer. They compare the measured spectral curve against the target curve. The colorist's expertise is in knowing that adding a touch of this pigment will fix the dip at five hundred fifty nanometers without throwing off the peak at six hundred.
It's like audio equalization, but for light.
That's exactly the right analogy. You're shaping a spectral curve the way an audio engineer shapes a frequency response.
The Paris Olympics last year gave us a real-world case study of this going sideways.
The twenty twenty-four Paris Olympics logo. The official blue was specified in Pantone, but when the fabric manufacturer for the uniforms tried to hit that blue on polyester — which takes dye completely differently than coated paper — they couldn't do it with the standard recipe. They had to develop a completely different pigment set that produced the same perceived color on fabric. Required a full spectral re-match, weeks of work, and I'm sure a very tense series of emails.
The Coca-Cola red problem, too.
Coca-Cola's red is Pantone four eighty four C. On a printed can, it's iconic. But on a backlit digital billboard, that same specification mapped to sRGB values looks different — usually more orange, sometimes washed out depending on the display technology. The company now maintains entirely separate color specifications for print, for signage, for digital, and for textile merchandise. Same brand, same "red," four different official numbers.
This is the thing that I think would surprise most people outside the industry. You assume a brand's color is one thing. But Coca-Cola red is actually a family of colors, each tuned to a specific medium, and they're all slightly different because they have to be.
That family is managed by a color standards team whose entire job is to make sure that when you see a Coke ad on your phone and then see a Coke can in a vending machine, the red feels consistent even though the actual spectral reflectance is completely different. It's perceptual consistency, not physical consistency.
The answer to "is there a single authoritative color scale" is a resounding no. It's a negotiation between physics, perception, and commerce.
The tools are evolving. There's been a significant shift in the last few years toward spectral data as the universal intermediary. Instead of saying "this is Pantone two eight six C," you say "here's the spectral reflectance curve." Every system can map to that. Pantone can give you their closest match. RAL can give you theirs. Your display can render its best approximation within the sRGB gamut. The spectral curve becomes the ground truth.
Which opens the door for open-source alternatives to break the proprietary lock-in.
The ColorHug project is an open-source colorimeter — not a spectrophotometer, but a solid colorimeter — that costs a fraction of what commercial devices run. And there are spectral databases like the Munsell Color Science Lab's datasets that provide reference measurements for thousands of colors without requiring a Pantone license. We're not at the point where Pantone's dominance is seriously threatened, but the cracks are forming.
The twenty twenty-four revision of the US Flag Code is a signal here. I saw they explicitly referenced the CIE Lab color space for the first time, moving away from the earlier Pantone-only specification.
That's a big deal. The US government is saying, in effect, "the authoritative specification for our flag is now a device-independent color space, not a proprietary ink-mixing system." It's a move toward spectral thinking, even if Lab is still a tristimulus space rather than full spectral data.
Let me try to synthesize what we've covered, because there's a lot here and I want to make sure we deliver something actionable. If you're specifying a color for a physical product, what are the three things you absolutely need to get right?
First, never rely on a single system. If your primary specification is a Pantone number, also capture a spectral measurement and note the illuminant conditions. If you're using RAL, also get the closest Pantone equivalent and the NCS notation if you're working in Europe. Redundancy saves you when the substrate changes or the lighting changes.
Second, for digital-to-physical workflows, invest in a spectrophotometer, not just a colorimeter. The colorimeter will tell you if two samples match under one light. The spectrophotometer will tell you if they match under any light. And learn to read spectral reflectance curves. The Delta E two thousand formula is your friend — it quantifies the perceptual difference between two colors. Anything under one point zero is imperceptible to the human eye.
Third, when dealing with national flags or government specifications, always check the official source. Don't assume the Wikipedia hex code is correct. Many flags have multiple official versions depending on the medium — fabric, paint, digital display, embroidery. The UN flag has different Pantone numbers for print versus digital. The Canadian flag has at least four standards. Go to the source document, not the fan wiki.
If you're a designer who's been treating Pantone as the universal truth, it's worth sitting with the discomfort of realizing it's a commercial product optimized for one very specific workflow — offset printing on coated paper. It's a brilliant product, it solved a real problem, but it's not physics.
The human eye can distinguish about ten million colors. Most colorimeters can measure about sixteen point seven million — twenty-four-bit color. That sounds like plenty, but it's only covering the gamut of a standard display. Real-world colors, especially saturated pigments and metallics, often fall outside that range. A spectrophotometer captures the full spectral reality.
There's a fun fact buried in there. The sixteen point seven million colors of twenty-four-bit color — that's not some fundamental limit of human vision. It's an engineering artifact of eight bits per channel. We built our digital color infrastructure around the constraints of nineteen-nineties computing, and now we're stuck with it.
We're bumping up against those constraints in weird ways. The latest generation of OLED displays can produce colors that fall outside the sRGB gamut entirely — deeper reds, more saturated greens. But most software doesn't know what to do with those colors because the entire pipeline assumes sRGB. You've got hardware capable of displaying colors that the software can't describe.
Where does this leave us? Color is never just color. It's a negotiation between physics, perception, and commerce. The standards we choose reflect what we value. If you value consistency across a global supply chain, you might choose RAL's spectral approach. If you value design intuition and ease of communication with printers, you might choose Pantone. If you value how a space feels to the humans in it, you might choose NCS.
The next frontier is going to make all of this look simple. We're seeing the emergence of dynamic color — materials that change color based on temperature, light, or electrical stimulus. Thermochromic pigments, photochromic dyes, electrochromic surfaces. How do you specify a color that doesn't stay still? What's the Pantone number for "blue, but only when it's cold"?
The philosophical question underneath all of this is that color isn't a property of objects. It's a relationship between a light source, a surface, and an observer. The standards we build are attempts to freeze one moment of that relationship and call it truth. And they work — until you change the light, or the material, or the observer.
As augmented reality and digital twins become more common, I think we'll see a convergence toward spectral data as the universal color language. A digital twin of a factory needs to render colors accurately under simulated lighting conditions, and the only way to do that is with full spectral information. The proprietary systems will fight it — there's too much money in selling swatch books — but the physics is on the side of open spectral data.
It's the MP3 moment for color. Once the spectral data is digitized and shareable, the business model of selling physical swatch books starts to look precarious.
Though I'll say this in Pantone's defense — there's still no substitute for a physical swatch when you're doing critical color work. Every display is calibrated differently. Holding the actual printed swatch next to the actual product under the actual lighting conditions that matter for your application — that's still the final arbiter. The technology augments that judgment, it doesn't replace it.
I think that's the thing that humbles people who are new to color work. You can spend fifty thousand dollars on spectrophotometers and calibrated displays, and then you take the product outside and hold it next to the swatch in natural daylight and realize your eyes are still the most sensitive instrument in the room.
The human visual system is astonishing. It can detect color differences that push the limits of our best instruments, and it does it instantly, without calibration, across an enormous range of lighting conditions. The tools are there to extend that capability, to make it communicable and verifiable across distance and time. But they don't replace the eye. They serve it.
The colorimeter is a useful tool, but it's not the whole answer. The whole answer is: know your substrate, know your illuminant, cross-reference your standards, and when it really matters, get a spectrophotometer and someone who knows how to read the curves.
Always, always check the official flag specifications before you print ten thousand units.
Words to live by.
Now: Hilbert's daily fun fact.
Hilbert: In the nineteen twenties, Icelandic geologist Guðmundur Einarsson proposed that the mineral troilite — an iron sulfide found in meteorites — was proof that volcanic basalt on Iceland's Snæfellsnes peninsula originated from lunar ejecta. The theory enjoyed mainstream acceptance for nearly a decade before spectroscopic analysis revealed the troilite was terrestrial contamination from a nearby iron smelter.
Iceland was briefly made of the moon because someone didn't clean their samples.
The moon of all places. Hilbert, did you plan that?
Hilbert: I neither confirm nor deny thematic alignment in the fact file selection process.
Of course you don't. This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop for keeping the lights on and the fact files weird. If you enjoyed this episode, leave us a review wherever you get your podcasts — it genuinely helps other curious humans find the show. We're back next week with another prompt. Until then, check your illuminant.
