Comparative Judgement (CJ) is traditionally and primarily used to compare written texts. In this study we explored whether we could extend its use to comparing audio files. We used GCSE Music portfolios which contained a mix of audio recordings, musical scores and text documents. Fifteen judges completed two exercises: one comparing musical compositions and one comparing musical performances. For each exercise, each judge compared 80 pairs of portfolios. Once judges had finished both exercises, they completed a questionnaire about their views and experiences of the method. Here, we present the judges' perceptions of using CJ in this context with reference to the Dimensions of judge decision-making model (Leech & Chambers 2022). We also compare the findings to those from text-based CJ studies.