Color can be created by many different methods. The main ones are additive and subtractive. The additive method combines colored lights, such as the red, green and blue phosphors on the screen of a computer monitor or television set.

The subtractive method uses pigments or inks such as in paintings and inkjet prints to subtract certain colors from white light and reflect the remaining colors to the eye.

CCD and CMOS chips are really only monochrome sensors. That means they see in only black and white and shades of gray. But through a neat combination of color theory, human perception, and computer processing power, they can produce color images in one shot.

Mouse Over

Hold your mouse cursor over the image to see a comparison between a Monochrome Color Filter Array (CFA) image and a Color Bayer-Interpolated Image. The CFA image may look like a normal black and white image at this image scale, but all of the information needed to create color is contained in the monochrome image.

Tri-Color

The traditional way to create color from a monochrome sensor is to shoot three separate exposures, one each through a red, green and blue filter. These three black and white filtered images are then combined into an RGB color image. This method of creating color was invented by James Clerk-Maxwell in 1861 but not perfected for many years. Even today this is the way color photographs are created with many high-end astronomical CCD cameras.

The Bayer Array

Bayer Filter Pattern

A Bayer array consists of a grid, or mosaic, where each row is made of alternating red and green, or blue and green filter colors. This produces a final mosaic with twice as many green pixels as either red or blue. Bayer found that because human perception was more sensitive to luminance (brightness) information in the green portion of the spectrum, increased luminance resolution by using more green pixels meant more perceived sharpness in the final image.

Color for each individual pixel is created by examining the color of the pixels around it and then processing this information through a sophisticated mathematical algorithm.

As you can see in the example at the top of this page, the Bayer interpolation algorithm does a remarkably good job of creating accurate color from the black and white original.

Mouse Over

Hold your mouse cursor over the image to see a comparison between a Monochrome Color Filter Array (CFA) image and a Color Bayer-Interpolated Image. These images have been enlarged 500 percent so that the individual pixels are visible.

Bayer Array Resolution

In re-constructing the color present in the original scene, the red, green and blue filtered black and white information from each individual pixel is compared to the information from its neighbors. Because color information from all three channels is present in each 4x4 pixel group, color for each individual pixel can be created through interpolation. The red, green and blue filters used over the individual pixels also have large regions of overlap in their spectral sensitivity. For example, some red and green light also gets into the blue pixel information. This means that the luminance resolution from a Bayer sensor array is correlated almost 1:1 with the pixel resolution, but the color information has less resolution. This is usually not a problem as human perception is less sensitive to color resolution.