The paper presents a new approach to the construction of test signal generators and test images, for colorimetric evaluation of the quality of functioning of multimedia channels of speech, military, and special purpose. The approach is to use adaptive properties of human vision to determine the number of colors for evaluation in certain adapting parameters. The main criterion for determining the number of colors is the color cut-off threshold. It is proposed to use color sorting capabilities to obtain frequency estimation parameters through a path guiding channel based on transition and ramp signal forms. Empirical expressions are given that allow a priori to estimate the magnitude of the change in the test signal and to model the possible magnitude of the effect on color rendering