# An Optimum 16 Color Palette

Assignment: Design a palette of 16 colors. Take into consideration a variety of tasks: search, sort, legibility, reproduction, and expressiveness. Rationalize your choice.

## Selection Rationale

In order to maximize the use of the color palette for classification, I started by selecting the colors "by name" using the eleven colors that Boynton says are almost never confused: black, gray, white, red, blue, green, yellow, orange, brown, purple, and pink. To these, I added another gray (giving four achromatic colors). The remaining four colors I selected (still only by name) as follows:

• To provide for better skin tones, and also to provide another very desaturated color with a different hue than pink, I added Tan.
• In order to provide a range of values, I added a light blue and a light green (pink already provides a very light red.)
• The remaining color I assigned to cyan, mainly cause I like blue-greens.

When it came to actually assigning colors to names, I attempted to minimize saturation differences between colors. Hue is well sampled, at roughly eight points, and lightness has five values (the two extremes are achromatic.) The colors are grouped as follows

### By constant hue

• Achromatic - Black, Dk. Gray, Lt. Gray, White
• Blue, Lt. Blue
• Cyan
• Green, Lt. Green
• Yellow
• Brown, Orange, Tan
• Red
• Purple, Pink

### By constant value

The value of the lower group was arbitrarily set. Light blue determined the value of it's group of colors. Yellow, tan and pink all "wanted" to be brighter.
• 0 % - Black
• 50 % - Dk. Gray, Red, Blue, Green, Brown, Purple
• 80 % - Lt. Gray, Lt. Blue, Lt. Green, Orange, Cyan
• 90 % - Tan, Pink, Yellow
• 100% - White

### By constant saturation

It is impossible to obtain a saturated light blue, therefore it and light green form their own little group.

• 0 % - Black, Dk. Gray, Lt. Gray, White
• 20 % - Pink, Tan
• 40 % - Lt. Green, Lt. Blue
• 80 % - Red, Blue, Green, Brown, Purple, Cyan, Orange, Yellow

A listing of the color palette is available.

## Search and Sort

This palette has a large diversity in hue. Since hue is the color property most useful for searching, it seems likely that this palette will be good for searching. A few of the worst case examples (each showing either a slight hue or value difference) are shown here:

## Legibility

The value groups listed above provide roughly 63 combinations of text and background (assuming light text on dark background) that should be easily legible. Here are some examples:

## Reproduction in Print

The colors were adjusted to provide good results when printing (yellow was moved to a higher brightness, and cyan was reduced in brightness). The colors with very low saturation (pink and tan) don't reproduce well.

## Reproduction of Real Images

The chosen palette should perform well with real images. In particular, special attention to skin tones is provided by two highly desatured colors (pink and tan) which provide good skin tone rendition for lighter skin, while darker skin tones are handled well by a brown.

Here's a test image (Card) shown using the above palette. In both it and the 16 color images below, Photoshop's error diffusion algorithm was used. For comparison, here are:

## Expressiveness

This color palette is capable of expressiveness, as shown by the following color pairs:

### Some Quiet Color Pairs

These color pairs generally differ only in hue, and are analogous colors.

### Finally, Strong Color Pairs

These color pairs generally differ only in value and saturation.

## References

Boy89
Robert M. Boynton. Eleven Colors That Are Almost Never Confused. In SPIE Proc. #1077 - Human Vision, Visual Processing, and Digital Display, pages 322-332, Bellingham, WA, 1989. SPIE.