Dithering is an imaging method used to add additional levels of shading where none actually exist. In the dark ages of computer graphics, where everything was either black or white, dithering was implemented to simulate shading. Several different forms of dithering was developed, and are often still available today. The most popular has been the summation dither, as it tends to offer the most natural results.
Dithering uses various noise patterns to blend between values. As computer graphics moved from black and white to 256 shades of gray, on through 256 shades of color, and even today, with 24 bit color and beyond, dithering is still useful. If you are asking yourself, “why?” the answer is because there is no color system on a computer that can mimic the ability of the human eye to see the vast amount of colors that we perceive, and at infinite resolution. In addition, color systems are not linear in response. If you were to view a color system like RGB as a curve, you would see that 24 bit RGB favors brighter colors over darker ones. As a result, even though you may have 24 bit color, you will perceive bands (“mach bands”) of color in the shadows. Fortunately, you can use dithering to hide these bands, and the result is very effective. And, because the dithering appears as noise, we accept the result since the look mimics film grain.