A quantitative study on color harmony

Abstract

In color vision, colors appear as interrelated visual sensations. Color harmony is about whether colors working together may produce a pleasing effect. It is a result of interaction between the external world and human's psyche. Apparently, it is a multidisciplinary research field, including perceptual and cognitive psychology, color vision, aesthetics, fashion, etc. This investigation examined what colors working together can invoke a harmonious feeling, how the color harmony is predicted, what the inner causation of color harmony is, and how color harmony models can be used in application. To transfer real-life situation into laboratory, color patches or chips were used to simulate object and/or background colors in this investigation. Two kinds of the most general arrangement of two colors, one of which was two colors laid side by side (SS pattern), and the other was two colors in center-background relationship (CB pattern), were designed for two-color harmony evaluation. In order to estimate three-color harmony, two kinds of three-color arrangements, one of which was three colors laid side by side on a grey background (SSS pattern), and the other was two colors laid side by side on a chromatic background (SSB pattern), were devised. Rating scale method was used to evaluate preference levels of single colors and harmony levels of color patterns. Twenty Chinese students participated in this aesthetic evaluation. By using multiple regression analysis, models of SS color pattern and CB color pattern were developed. 72.6% of total variance of SS patterns and 66.3% of total variance of CB patterns can be explained by their corresponding models. Component color relationships and single color preference are the two main categories of factors affecting color harmony of both color patterns. Characteristics of each single color in the pattern are considered to be less important to the harmony of a color pattern. Hue difference is the primary influential factor on color harmony of SS patterns, while lightness difference is of the most importance for CB patterns. A positive relationship between color harmony and total lightness was found in both kinds of patterns. In view of the strong positive dependence of single color preference on lightness, it is concluded that high harmonious levels of patterns with high total lightness might attribute to both their high total lightness and high single color preference. For both kinds of three-color patterns, a positive correlation was found between harmony of three-color patterns and that of related two-color patterns. However, although it is feasible for SSS patterns to predict their harmony values from those of three related two-color patterns, it is difficult for SSB patterns possibly due to their complex structures. By combining these color harmony results with sensation and perception theory, it is deduced that color harmony is a pleasing feeling evoked by color patterns which consist of preferred single colors and can satisfy people's expectation. Since the single color preference is related to the cultural background of observers, harmonious levels of color patterns may be different for peoples with different cultural background. Finally, to apply the color harmony models to practical use, a flexible and general method to establish harmonious color patterns was developed which was applicable to all kinds of color pattern design. By progressively setting the constraints such as the function required by the situation, the emotion to be expressed, the influence of prevailing color, on color patterns, the most harmonious color combinations could be determined based on the color harmony models. In conclusion, this investigation demonstrates that color harmony is a stable feeling and can be objectively quantified by aesthetic experiment. It provides a framework for further research on color harmony.