Is My Blue Your Blue?

People have different names for the colors they see. Language can affect how we memorize and name colors. This is a color naming test designed to measure your personal blue-green boundary.

Test validity

Color perception is tricky to measure--vision scientists use specialized calibrated equipment to measure color perception. Graphic designers use physical color cards, such as those made by Pantone, so that they can communicate colors unambiguously. Here we use your monitor or phone to test how you categorize colors, which is far from perfect, since your calibration may differ from mine.

The validity of the inference is limited by the calibration of your monitor, ambient lighting, and filters such as night mode. Despite these limitations, the results should have good test-retest reliability on the same device, in the same ambient light, which you can verify by taking the test multiple times. If you want to compare your results with friends, use the same device in the same ambient light.

Getting outlier results doesn't mean there's anything wrong with your vision. It might mean you have an idiosyncratic way of naming colors, or that your monitor and lighting is unusual.

Technical details

The test asks you to categorize colors sequentially. Colors are often represented in HSL (hue, saturation, lightness) color space. Hue 120 is green, and hue 240 is blue. The test focuses on blue-green hues between 150 and 210. On the web, HSL coordinates are translated to sRGB color space, the standard color space of the web, which is not perceptually uniform. These sRGB values are translated nonlinearly to your monitor through a gamma curve.

This is equivalent to a logistic regression model. The test uses a maximum-a-posteriori (MAP) estimation algorithm (specifically, a second order Newton method implemented in pure JS, no calls to a backend) to fit the sigmoid curve to your responses, with a vague prior on the scale and offset parameters. It uses the fitted curve to determine which color will be presented next. It tries to be smart about where it samples new points, focusing on regions where you're predicted to be intermediately confident in your responses. To improve the validity of the results, it randomizes which points it samples, and uses a noise mask to mitigate visual adaptation.

It's a curve fit, not a binary search. In theory, if you feel like you're guessing in the middle shades, or even guessing incorrectly, that should be fine. If you're inconsistent in the middle, the curve fit should be able to recover, although your estimated threshold will have larger error bars.

Results

In early experiments, we found that people's responses cluster around 175, which coincidentally is the same as the named HTML color turquoise. This is interesting, because the nominal boundary between blue and green is at 180, the named HTML color cyan. That means most people's boundaries are shifted toward saying that cyan is blue.