“From Sound to Shade”: How Emotion and Music Genre Guide Crossmodal Color Selection

ABSTRACT This study investigates music–color crossmodal correspondences, focusing on how target‐emotion condition and musical genre jointly shape people's choice of color and perceived coherence, defined here as the subjective strength of the match, or correspondence, between a musical excerpt and a chosen color. In a within‐subjects experimental design, 278 participants evaluated 15 music excerpts crossing Joy, Sadness, and Nostalgia with Classical, Pop, Rock, Jazz, and Electronic genres, selecting colors in CIELAB space and rating coherence and emotional intensity. Robust Emotion × Genre interactions were demonstrated for coherence and color dimensions. In particular, luminance (L*) emerged as the primary visual cue, with emotion explaining substantial variance with participants choosing to map bright colors to joy and darker, more saturated colors, to sadness. Chroma, but not hue, consistently tracked emotional meaning, while hue showed greater genre dependence. Emotional intensity added minimal predictive value beyond the experimental conditions, although this effect should be interpreted with caution, given the relatively limited variability in intensity ratings across conditions; the present analyses did not directly compare valence and arousal with categorical emotion variables. Mediation tests indicated that L* functions as a suppressor in the emotion–coherence relationship. These findings are consistent with an emotion‐related account of music—color correspondence while demonstrating that musical genre constrains multisensory mapping, with luminance emerging as a perceptual dimension that consistently covaries with the target‐emotion conditions across the sampled stimuli. Because participant‐reported experience was not shown to mediate these effects, the results should not be taken as a direct test of emotional mediation in the stronger causal or mechanistic sense. As with any single‐exemplar‐per‐cell design, these effects are most directly interpretable at the level of the sampled stimuli; the large, well‐powered participant sample strengthens confidence in the reported patterns for these excerpts specifically, and extending the design to multiple exemplars per cell is a natural next step for establishing generality across the broader emotion and genre categories.