Mapping model from chord to color

Abstract

As an art, music has been used by way of binding with some other modality or media, e.g. color and pictures. In this study, the multimodal perceptual cross representation between chord and color has been focused on. Most of cultures have some musical songs as well as language. Musical chord representation of human is not clear. The some musical representation was reported for some area, such as scaling. And some methods for the chord feeling were proposed. Mapping between language and chord could depend on their own culture. There are some evidences that indicate the interactive relationship between language and color categorization. The color concepts depend on language or culture but color discrimination would be independent of language. This indicates that it is necessary to use the non-verbal way in order to elucidate about color-chord representation independent of cultures. Our purpose in this study is to elucidate the relationship between chord and color and to validate the some chord features proposed in past musical studies. In this study, we examined chord-color mapping experiments. The mapped color data was analyzed in the HLS color space, consisting of hue, saturation and luminance. To elucidate the relationship between chord and color representation, we applied the statistical model analysis from chord features to color feature and model selection method by the AIC. As the result, we examined the relationship from musical stimuli to color mapping and showed that the relationship exists for color and chords by the linear model analysis. From the nonverbal mapping experiment in this study, we showed that the chord feature values have an important role about musical representation of chord perception. This indicates that the musical chord features was validated as the chord representation.