The Science of Color Perception: How Your Brain Sees Color

The Hardware: Cone Cells in Your Retina

Human color vision begins in the retina, a thin layer of neural tissue at the back of the eye. The retina contains two types of photoreceptor cells: rods (about 120 million, responsible for low-light vision) and cones (about 6–7 million, responsible for color). Cones are concentrated in a small area called the fovea, which is why you see color best when looking directly at something.

There are three types of cone cells, each containing a different photopigment (opsin) that responds most strongly to a specific range of wavelengths:

Notice that no cone type peaks at "red." L-cones peak in the yellow-green range but have a long tail extending into the red wavelengths. Our perception of red comes from the difference in response between L-cones and M-cones, not from a dedicated "red detector." This is a crucial point that many people get wrong.

The distribution of cones is also uneven. L-cones and M-cones together make up about 90–95% of all cones, with S-cones comprising only 5–10%. This is why we are more sensitive to differences in the red-green range than in the blue range — a fact that has direct implications for color wheel design and digital color models.

Trichromatic Theory: Young-Helmholtz (1802–1867)

The first major theory of color vision was proposed by the English polymath Thomas Young in 1802 and later refined by the German physicist Hermann von Helmholtz in the 1850s–1860s. The Young-Helmholtz trichromatic theory states that all perceived colors can be produced by varying the stimulation levels of three types of receptors. Any color we see is encoded as a triplet of activation values: (S, M, L).

This theory elegantly explains several phenomena. It explains why we can reproduce any visible color by mixing just three primary lights (the basis of RGB displays). It explains why people with only two types of functioning cones (dichromats) have reduced color discrimination. And it predicts the existence of metamers — physically different light spectra that produce identical (S, M, L) responses and therefore look the same to us.

The theory was confirmed in the 1960s when researchers George Wald, Paul Brown, and Edward MacNichol used microspectrophotometry to measure the absorption spectra of individual cone cells. Wald received the Nobel Prize in Physiology or Medicine in 1967 for this work.

Opponent-Process Theory: Ewald Hering (1878)

Trichromatic theory could not explain everything. Why can we imagine a "reddish-yellow" (orange) but not a "reddish-green"? Why do afterimages appear in complementary colors? The German physiologist Ewald Hering proposed an alternative in 1878: the opponent-process theory.

Hering argued that color vision operates on three opponent channels:

• Red vs. Green — signals cannot be simultaneously red and green
• Blue vs. Yellow — signals cannot be simultaneously blue and yellow
• Black vs. White — the achromatic (lightness) channel

For decades, the two theories were seen as rivals. It was not until the 1950s–1960s that researchers, including Leo Hurvich and Dorothea Jameson (1957), showed that both theories are correct — they simply operate at different stages. Trichromatic processing happens at the receptor level (cones), while opponent processing happens in the retinal ganglion cells and lateral geniculate nucleus (LGN) of the brain. The cone signals are combined as follows:

This dual-process model — trichromatic at the receptor level, opponent-process at the neural level — is the current scientific consensus and forms the biological basis for color spaces like CIE L*a*b*, which maps directly onto the lightness, red-green, and blue-yellow opponent channels.

The Dress Illusion (2015): When the Internet Broke Over Color

On February 26, 2015, a photograph of a dress went viral on Tumblr, sparking one of the largest perceptual science events in history. Some people saw the dress as blue and black; others saw it as white and gold. Within 48 hours, the image had been viewed over 10 million times.

The phenomenon was studied by researchers including Pascal Wallisch (New York University), Karl Gegenfurtner (Justus Liebig University Giessen), and a team at MIT. Their research, published in journals including Current Biology (Lafer-Sousa et al., 2015) and Journal of Vision (Wallisch, 2017), revealed that the disagreement was caused by chromatic adaptation — the brain's attempt to discount the illumination and determine the "true" color of the object.

People who assumed the dress was lit by warm, yellowish light (like sunlight) subtracted yellow from the image and saw blue-black. People who assumed cold, bluish light (like shade) subtracted blue and saw white-gold. Wallisch's study found that early risers (exposed to more daylight) were more likely to see white-gold, while night owls were more likely to see blue-black. The dress demonstrated that color perception is not passive measurement but active, assumption-driven neural computation.

Color Constancy: Your Brain's White Balance

The dress illusion is an extreme example of color constancy, a perceptual mechanism studied extensively by Edwin Land (inventor of Polaroid) in his Retinex theory (1977). Color constancy is your brain's ability to recognize that a red apple is still red whether you see it under fluorescent office lighting, warm candlelight, or outdoor sunlight — even though the wavelengths reaching your eye are vastly different in each case.

This is why designers need to test colors in multiple lighting conditions, and why a color converter that shows you the exact numerical values of a color is more reliable than your eyes alone.

The Mantis Shrimp: 16 Color Receptors

Humans have three types of color receptors. The mantis shrimp (Stomatopoda) has sixteen. When this fact went viral (partly thanks to a famous comic by The Oatmeal), many people assumed the mantis shrimp sees millions more colors than we do. The reality, revealed by research from Hanne Thoen and Justin Marshall at the University of Queensland (published in Science, 2014), is more nuanced and more interesting.

Thoen et al. found that mantis shrimp are actually worse than humans at discriminating between similar colors. Their 16 receptor types appear to function as a "barcode" system rather than a mixing system. Instead of combining signals from a few receptors (as we do), they bin wavelengths into 16 categories for fast, efficient recognition. It is a speed-optimized system, not a resolution-optimized one — useful for a predatory animal that needs to identify prey and rivals in milliseconds in the visually complex coral reef environment.

Tetrachromacy: The Humans Who See Extra Colors

While most humans are trichromats (three cone types), a small number of women may be functional tetrachromats — possessing four types of cone cells. Research by Gabriele Jordan at Newcastle University, published in the Journal of Vision (2010), identified at least one confirmed tetrachromat who could distinguish colors that appeared identical to trichromats. Tetrachromacy is theoretically possible in women because the genes for the M and L cone opsins are located on the X chromosome; having two X chromosomes can result in a variant fourth cone type.

Estimates suggest that up to 12% of women may carry the genetic basis for tetrachromacy, but only a small fraction develop the neural pathways to actually use the fourth channel. For those who do, the color world is richer — they may perceive up to 100 million colors, compared to the roughly 1 million visible to typical trichromats.

The Visible Spectrum by the Numbers

The electromagnetic spectrum spans wavelengths from picometers (gamma rays) to kilometers (radio waves). The visible portion is a remarkably narrow band:

This means the entire rainbow of human color experience occupies a range from 380 nm to 750 nm — less than one octave of the electromagnetic spectrum. Everything you see when you browse color palettes or create gradients lives within this narrow band.

Conclusion

Color perception is not a camera passively recording wavelengths. It is a multi-stage computational process: trichromatic encoding at the cone level, opponent-process transformation in the retina and LGN, and high-level contextual interpretation in the visual cortex. Your brain makes assumptions, fills in gaps, discounts illumination, and constructs an experience of color that serves survival rather than physical accuracy.

Understanding this science makes you a better designer. It explains why colors shift in different contexts, why some people see colors differently, and why perceptually uniform color spaces (like L*a*b*) were developed to match how we actually see, rather than how screens emit. Every tool on Colors Hunter is built on these principles — because working with color means working with human perception.