Avoid very dark colormap regions on white backgrounds

Avoid very dark low-end colors when a quantitative colormap is shown on a white background. For example, lighten or remove the black end of greys, magma, or plasma on a white heatmap or legend; those dark tails produced large error spikes.

Very dark colors on white can look more separated from the background than from each other, so readers struggle to judge which dark shade is closer to a reference.

Mechanism: High contrast against white interferes with discrimination among nearby dark shades, especially near the low end of the scale.

Evidence: The study found dramatic error increases in the dark regions of greys, magma, and plasma on a white background, and these failures were much worse than predicted by LAB or UCS color-distance models (Liu & Heer, 2018).

Notes: The paper warns that an analogous problem may also occur for very light colors on dark backgrounds, but that case was not tested.

• User Goal: Compare nearby low values accurately.
• Task: Judge relative distance between dark shades on a quantitative scale.
• Data: Ordered quantitative values encoded with a continuous colormap.
• Chart Setting: A white or high-luminance background with the dark end of the palette visible.
• Audience: Viewers reading the low end of the scale directly.
• Success Criterion: Accurate low-end comparisons rather than visual drama at the darkest values.

Break it when: The chart is not shown on a white or other high-luminance background. Why: The measured degradation was tied to dark colors set against white, and the paper does not claim the same magnitude of failure for other backgrounds.

Sacrifice: You give up the dramatic black endpoint used by some palettes. Risk: Keeping the black endpoint for style can make nearby low values hard to compare. Mitigation: Start the ramp above black or choose a palette whose low end is not extremely dark when the background stays white.

Mistake: Trusting perceptual color-distance numbers alone for the darkest colors on a white background. Why it fails: The study found much worse discrimination there than the color models predicted.

Failure Sign: The darkest two or three steps of the scale look hard to order on the actual white chart. Quick Check: Inspect nearby low-end values on the chart background, not only in a palette editor. Stronger Test: Compare a dark-end palette with a lighter-start alternative on the same white chart and keep the version with fewer low-end errors.

• Remove or lighten the darkest endpoint of the colormap.
• Switch to a palette whose low end does not collapse into near-black on white.
• If the dark palette must stay, change the chart background so the low end is not shown against white.