This class defines reduction options for scale reduction. Scale reduction can be performed on an image whose color space is anything but indexed. The output image will always have the same color space as the input image. It works by linearly scaling the component values for each pixel to the desired output depths. For example, you could reduce a 24bpp (8-8-8) RGB image to 16bpp (5-6-5) RGB. Simply specify the desired output channel depths (and color space) in the parameters for the Reduce method.

Note: You can set DontScalePixels option to true to avoid scaling the pixel values. In this case, the component values of each pixel will be preserved (within the limits of precision) even though their effective ranges may be different.

Reference

ImGearScaleReductionOptions Members
ImageGear.Processing Namespace