Science of Warm and Cool

 

In painting, color temperature is such an ambiguous concept. For example: orange vs yellow, which one is warmer? Ultramarine vs phthalo blue, which one is cooler? Artists do not want to quantify their color temperature. Their argument is: color is so relative. We only see a color warmer or cooler with respect to another color. This fuzzy ambiguity bothers my engineering mind. So I went to the field of photography. Lo and behold! I found their concept of color temperature is so crispy clear. They quantified the color temperature using K or Kelvin. The "warm" colors have lower K, while "cool" ones higher K. Therefore: orange is warmer than yellow, violet is cooler than turquoise (so ultra is cooler than phthalo). I found a photographer's temperature scale online, and measured 11 color samples. I plot those samples on the CIELuv (pls ignore this nomenclature except the color scientists) color wheel. I was shocked by what I have found. Ah-ha!! Now I know. (1) For very warm colors (red to yellow, or 1 to 4) the temperature varies with the HUE change. Similarly for very cool colors (turquoise to violet, or 8 to 11). This finding reminds me the "prismatic palette" some artists talked about. They follow the rainbow when they paint. (2) For the moderate warm and cool colors (from yellow to blue, or 4 to 8), the temperature varies with the CHROMA change, then flip the hue at the center of the color wheel to the complementary. I feel this understanding can help us to paint better: when you deal with very warm colors like sunset, you can change temperature by hue shifting. However when you deal with grey and subtle colors, you change temperature by chroma variation. Voila!