Virtuosi impress audiences with their musical expressivity and with their theatrical flair. How do listeners use this auditory and visual information to judge performance quality? Both musicians and laypeople report a belief that sound should trump sight in the judgment of music performance, but surprisingly, their actual judgments reflect the opposite pattern. In a recent study, when presented with 6-second videos of music competition performers, listeners accurately guessed the winners only when the videos were muted. Here, we successfully replicate this finding in a highly-powered sample but then demonstrate that the sight-over-sound effect holds only under limited conditions. When using different videos from comparable performances, in a forced-choice task, listeners' judgments were at or below chance. And when differences in performance quality were made clearer, listeners' judgments were most accurate when they could hear the music-without audio, performance was at chance. Sight therefore does not necessarily trump sound in the judgment of music performance.