Writing direct coloring algorithms

Direct coloring algorithms directly output a color instead of an index value. This is done by assigning a value to the #color predefined symbol instead of to #index.

To compute this color, you can use color expressions, color variables, and color arithmetic. The following arithmetic operations are available:

c1 + c2	Returns a color where each component is the sum of the respective components from c1 and c2. So, red(c1 + c2) is equal to red(c1) + red(c2).
c1 - c2	Subtracts the color components in c2 from the respective color components in c1. So, red(c1 - c2) is equal to red(c1) - red(c2).
*c f**	Multiplies each component of c with a float value. So, red(c * f) is equal to red(c) * f. Note: the float value must be at the righthand side of the * operator.
c / f	Divides each component of c by a float value. So, red(c / f) is equal to red(c) / f.

For example, to calculate the average of two colors, use (c1 + c2) / 2. Be aware of the fact that the alpha value is treated just like the other components. So, c / 2 will not only darken a color, it will also make it more transparent.

The following conversion functions are available:

There are some functions for blending and retrieving gradient colors:

There are also functions to reproduce all layer merge modes. They are called mergeX, where X stands for the name of the merge mode. See Merging functions.

You can use color parameters to let user specify a color. You can also use special user functions that allow a user to select a merge mode. See Parameter blocks.

Next: Global sections