A colored original to be reproduced is scanned point-by-point in each of the three primary colors. The resulting imaging signals are processed to generate a luminance signal and a pair of chrominance signals. The luminance signal is branched into a high-pass filter channel and a low-pass filter channel. The high-pass portion of the luminance signal is amplified in accordance with a sublinear characteristic function which causes signals of small amplitude to be amplified to a greater degree than signals of large amplitude. The low-pass portion of the luminance signal is modified per a non-linear characteristic gradation function. The thus-modified high-pass and low-pass portions of the luminance signal are added to yield an enhanced luminance signal. The enhanced luminance signal is divided by the original luminance signal and each of the chrominance signals is multiplied by the resulting quotient. This automatially compensates for the changes in color saturation which accompany changes in brightness. After the multiplication, the chrominance signals are amplified in accordance with a sublinear characteristic function which causes signals of small amplitude to be amplified to a greater degree than signals of large amplitude. This characteristic function further has a form such that no additional ampification takes place beyond a predetermined threshold limit.