In this project, we present the concept of backdrop visualizations - visualizations that support conversations without interrupting them. Our term comes from photography where backdrops form a frame without distracting from the main motif. Similarly, backdrop visualizations support the main task of discussion without actively drawing attention to them. This does not mean, however, that backdrop visualizations completely disappear into the ambient environment. They also work as look-up tools that sometimes become the center of attention. This means that such visualizations have to fulfill two roles - a passive one: as non-distracting but inspiring backgrounds, and an active role: as effective data look-up tools, and that they switch between those states at random intervals depending on the conversation.
arcs.fm is one specific example of a backdrop visualization for the domain of music talk. The visualization shows two personal music listening histories as a background for the conversation. From time to time, the conversational partners might use it for two active tasks that are central for backdrop visualizations: (1) for quick look-ups of longer and shorter periods of time (e.g., to find the name of a song), or (2) for exploration and highlighting interesting musical correlations which neither conversational partner is aware of (e.g., a shared liking for a specific artist).
Music is an appropriate topic for this exploration because it plays an important role not only in upholding one's self-image but also in determining one's perception by peers: adolescents frequently engage in discussions about music and even later in life, taste in music still retains an important function for identity management. A music listening history is an accurate reflection of this taste in music, but framing it and providing commentary is important. The power of music talk (we use the term following Frohlich et al.'s notion of photo-talk for similar discussions about photos, comes from the freedom to emphasize personally important aspects, providing a personal perspective on events, being able to quickly switch facets and the general to-and-fro between participants.
While people shine with these verbal activities, their memory is not flawless: even when discussing recent musical favorites they might forget about some of them and providing an accurate overall description of how one's taste in music developed is difficult. In addition to this frailty of memory, music talk only uncovers interesting similarities such as a shared favorite artist or knowledge of an uncommon genre by accident.
Using visualization in concert with automatically collected music listening histories can enhance these tasks.