We designed and implemented Musée to capture the novel experience of interpreting cover versions of music, which contain both familiar and unfamiliar musical components and are curated based on the user’s music-streaming history data. Musée is a tangible music player that enables users to explore and listen to professional or amateur covers of songs (via YouTube) in two categories: covers of songs from users’ most-liked artists and covers of users’ most-played songs. To investigate its potential value in situ, we conducted field trials of Musée in four households for 1 month. Findings showed that unfamiliar musical elements in cover music provided a sense of ‘freshness’ to past songs and helped the listener appreciate over-consumed music in new ways. In addition, restricting detailed information about cover songs that were playing helped users focus on the sound, thus priming them to infer and reflect on the original song and their memories associated with it. Our findings point to new insights for the design of interfaces that use historical personal data to expand users’ experience beyond solely revisiting prior tastes.