The "Geometry" of Colours

Summary

This video explores the science of color, diving into the history and geometry behind how we perceive it. From Aristotle to Schrödinger, thinkers have debated the nature of color, revealing that it’s a subjective experience created by our brain when light interacts with our eyes. The video explains the physical properties of light and color, color mixing, and how different color models like Newton's circle and Helmholtz's horseshoe diagram attempt to represent these perceptions. It also touches on psychological approaches to color, including how humans perceive distances between colors and how research continues to evolve with new models like okLab.

Takeaways

• 🎨 Colours are subjective sensations created by the brain in response to light; objects do not possess intrinsic colour.
• 🌈 The visible spectrum is a 1-dimensional space of pure wavelengths, ranging from red to violet, but mixtures of wavelengths create new colours with varying saturation.
• 🔺 Additive colour mixing (light) is different from subtractive mixing (pigments); the video focuses only on additive mixing.
• ⭕ Newton introduced the colour circle to visualize hue relationships and mixtures, adding purples which do not exist in the spectrum.
• 🔦 Adding brightness as a third dimension creates a full 3D colour space shaped like a cone, organizing hue, saturation, and luminosity.
• 🐎 Helmholtz corrected Newton’s model by distorting it to reflect perceptual and mixture-based realities, resulting in a horseshoe-shaped diagram that influenced the later XYZ space.
• 📊 The CIE 1931 XYZ colour space introduced a mathematical way to describe and mix colours, though screens (sRGB) only display a limited subset of these colours.
• 👁 Human colour perception depends on three types of cones; variations or deficiencies (colour blindness) reduce the dimensionality of perceived colour space.
• 🧠 Psychological and perceptual models (Goethe, Hering, Munsell, CIELAB, okLab) aim to represent colours in terms of how humans actually perceive differences, rather than purely physical properties.
• 🌀 Perception is non-linear (Weber-Fechner law), so uniform physical gradients do not look uniform; perceptually uniform spaces like Lab or okLab fix this.
• 🧬 Evolution shaped colour vision: early organisms saw only light vs dark, later blue–yellow, then red–green; some humans today may be tetrachromatic.
• 📈 MacAdam ellipses show that people distinguish colour differences unevenly across the spectrum, revealing distortions in traditional colour diagrams.
• 🖥 Colour models matter for accessibility, display design, and data visualization, especially for ensuring readability and colour-blind–friendly content.
• ✨ In 2025, researchers claimed to produce a completely new colour sensation, “olo”, by selectively stimulating medium cones with lasers, generating a colour outside the usual perceptual space.

Q & A

What is the main concept explored in the video?

The video explores the geometry of colors, investigating how colors are perceived, how color spaces are constructed, and the relationship between physical, psychological, and mathematical models of color.

Why do objects not have an intrinsic color?

Objects do not have intrinsic color because color is a subjective perception created by our brain when our eyes receive light. The color we see depends on the wavelengths of light an object reflects or absorbs.

How does sunlight decompose into different colors?

Sunlight can be decomposed using a glass prism. The waves interact with atoms in the prism, and each wavelength is deflected differently, producing the colors of the rainbow, which range from red to violet.

What is the difference between additive and subtractive color synthesis?

Additive color synthesis involves mixing different light waves (like on a computer screen) where each light adds to the mixture. Subtractive color synthesis, used in pigments like paints, works by absorbing light wavelengths rather than adding them.

What was Newton’s contribution to understanding color geometry?

Newton proposed a circular model of color, where the diagram closed on itself, with white at the center. He believed that mixing two lights would result in a color halfway between the two, and complementary colors would add up to white.

Why did Helmholtz modify Newton's color diagram?

Helmholtz modified Newton's diagram to address discrepancies. For example, he adjusted the curve between red and yellow to be flatter and placed indigo further from white to better reflect how our eyes perceive mixtures of colors.

What is the XYZ color space, and why was it created?

The XYZ color space was defined in 1931 by the International Commission on Illumination. It was designed as a more precise mathematical space to describe color mixtures and extend beyond Newton's model, accounting for different light intensities and color interactions.

What is the main limitation of current color displays, like monitors?

The main limitation is that monitors, which typically use RGB color mixing, cannot display all the colors that humans can perceive. This is due to the limitations of the screen's technology, which only covers a portion of the visible spectrum.

How does color blindness affect color perception?

Color blindness occurs when there are deficiencies in the retina or visual system, affecting the ability to perceive certain wavelengths. This results in a reduced color space, with some individuals perceiving only two or even one dimension of color.

What is the significance of the okLab color space?

The okLab color space, proposed in 2020, is a more precise model designed to represent how humans perceive color distances. It offers a more accurate geometric description of color perception, especially for screens and color pickers.