Understanding color reduction methods
When you decrease the color depth of an image, you must select a color reduction method. Depending on the color depth you select, two or three of these methods are available:
• Nearest Color — replaces the original color of a pixel with the color in the newly generated palette that is closest to the original RGB value. This method eliminates dithering and produces a high-contrast image. This method works well with most simple graphics.
• Error Diffusion — replaces the original color of a pixel with the most similar color in the palette, but spreads the discrepancy between the original and new colors to the surrounding pixels. As it replaces a color (working from the top left to the bottom right of the image), it adds the “error,” or discrepancy, to the next pixel, before selecting the most similar color. This method produces a natural-looking image and often works well for photos or complex graphics. With the Error Diffusion method, you select the Floyd-Steinberg, Burkes, or Stucki algorithm for the dithering pattern.
• Ordered Dither — adjusts adjacent pixels so that two colors give the illusion of a third color and intermingles pixels to produce patterns based on a known palette. You can use this method to give the image the appearance of containing more colors than it actually does. Images may appear to be composed of cross-hatches and dots and may have distinct light and dark areas.
When you decrease an image’s color depth, don’t be intimidated by the mathematical algorithms that make up the color reduction methods and the palette options. It’s recommended that you start with the default settings. If you like the result, try out several other combinations of settings. It take can take some experimentation to find the right options.