I was sitting with Dan Margulis’s book, the LAB color one, “Photoshop LAB Color: The Canyon Conundrum,” and I was about halfway through it and genuinely having my mind rearranged chapter by chapter, when my phone rang and it was one of my students. She was working at a high-end photography studio, real professional work, and she had this project she was excited about, a black-and-white series where they wanted to hold the eye color and render everything else monochromatic. You know, the eyes stay vivid, everything else goes grey. And she said it worked beautifully for brown eyes, decent results for green, but blue eyes were giving her nothing or worse than nothing, the eyes were actually rendering brown in the final image, and she wanted to know what was happening.
And I looked down at the book in my lap and the answer was sitting right there on the page I’d just been reading.
Here’s what I told her, and here’s what I want to tell you, because this is one of those things that sounds simple when you say it out loud but then you sit with it for a minute and it gets genuinely strange. The blue you see in someone’s blue eyes isn’t in their eyes. Not really. It’s a product of contrast with the surrounding skin tones, and when you remove that color context, when you go to black and white and strip away the warm pinks and peaches and siennas of the skin around the iris, the color perception collapses. The brain built that blue from context, and you just removed the context, so the brain has nothing to build from. And what’s left in the iris itself, the actual pigment, the actual physical reality sitting there in grayscale? Brown. Maybe grey. Not blue.
The technical name for this is simultaneous contrast, and Margulis walks through it in a way that I found almost unsettling, because once you see it you can’t unfocus your eyes and make it go away. The color you perceive in any given area is always, always being negotiated by the colors around it. Always. Your visual system doesn’t measure color independently the way a spectrophotometer does; it’s constantly comparing, contrasting, and inferring, and it is remarkably willing to tell you something is there that isn’t, or hide something that is, based entirely on what’s sitting next to it.
So a few days after that call, I went to find Carrie. She worked a couple of offices down from me and she had very pale blue eyes, the kind where the color seems almost transparent, very dramatic, you’d notice them immediately. And I asked her if she minded being a test case for something I was working on, and she was a good sport about it, and I took a photograph and brought it into Photoshop and isolated the iris, just the iris, and started changing the surrounding skin tones. And you know what happened. Same iris. Same pixels. Same actual color data sitting in that file. But as I shifted the skin tones warmer, cooler, lighter, darker, that iris read as blue, then shifted toward purple, then pushed into green territory. None of those colors existed in the iris. The iris hadn’t changed. What changed was the context, and the context was doing all the perceptual work, and the iris was just along for the ride.
Carrie thought it was pretty cool. I thought it was, honestly, a little bit of a philosophical problem.
Because here’s where this stops being just a fun Photoshop trick and starts mattering for how you work. If color perception is constructed rather than received, if your brain is actively building the color experience rather than passively recording what’s there, then every design decision you make about color is actually a decision about relationships, not about individual values. You think you’re choosing a blue. You’re actually choosing a blue-in-the-context-of-everything-else-on-the-screen, and those are not the same choice. The blue that reads clearly and crisply on a white background might go muddy and ambiguous on a warmer background, and it’s the same hex value, the same RGB numbers, the same color. Different perception entirely.
And this is where I want you to think about accessibility, because I think designers sometimes treat accessibility as a constraint, a set of rules you follow to check a box, and I’d push back on that pretty hard. The reason contrast ratios matter in accessibility guidelines isn’t bureaucratic, it’s perceptual. It’s rooted in exactly this phenomenon. Someone with low vision or color vision deficiency isn’t experiencing a lesser version of what you designed; they’re experiencing a different construction of it, one where the brain has less signal to work with and is doing even more inferential work than usual, and if you haven’t designed with genuine contrast, with relationships that carry meaning even when the perception varies, then what you made doesn’t work. Not aesthetically, not functionally. It just doesn’t work.
Now, I want to go back to Margulis for a second, because the simultaneous contrast stuff was only the beginning of what that book did to my head. I kept reading into the later chapters and things got genuinely strange, because he starts exploring colors that can mathematically exist in the LAB color space but cannot be perceived by human vision. A yellow that is as bright as white, that you can’t really see, that doesn’t exist in the human visible spectrum but mathematically exists and can be used in color correction. And I read that and I just kind of… stopped. Set the book down. Hadn’t come back to those chapters yet, if I’m being honest (and I’m being honest). That’s a level of abstraction I need a longer afternoon and a quieter brain for. But the early chapters, the simultaneous contrast material, that I’ve taught from ever since, because it reframes color not as a property of objects but as a product of relationships, and that reframe alone is worth the price of the book.
What I took away from all of it, from the student’s call and Carrie’s iris and Margulis and those impossible LAB colors sitting just outside human perception, is something that sounds obvious once you say it and yet is constantly violated in practice: what we see is not what is there. The brain is not a camera. It’s an interpreter, and it’s making judgment calls constantly, filling gaps, borrowing context, constructing a coherent experience out of wildly incomplete information, and it does this so seamlessly and so fast that we walk around completely convinced that what we’re seeing is just… reality. Objective. There.
And I think that humility, knowing that your eye is a negotiating instrument and not a measuring instrument, is one of the most useful things a designer can carry. Because it means you’re responsible for the context, not just the element. It means you can’t hand someone a color and say “here, this is blue” without also being responsible for everything surrounding that blue and what that surrounding is doing to the perception of it. You don’t get to opt out of the relationship.
Next time you look at someone with blue eyes, just hold that for a second. Your brain built that.
—
This connects to some of what I covered in the color theory post (that’s #88 if you want to find it) and the visual communication piece (#84), because this kind of perceptual reality underlies all of it. If you’re curious about simultaneous contrast and want to go deeper, Margulis’s book is genuinely worth your time, at least the first half, and I say that as someone who put down the second half and hasn’t picked it back up. Yet.
Keep Reading
- My Musical Taste Stopped Evolving in 1983 (and That Turned Into a Research Question)
- Color Is a Frequency. That Changes Everything.
- CRAP, But Make It Color
- Color Is Three Things at Once
Go deeper: The Neuroscience of Creativity · All Primers
