Listening With Machines: A shared approach

The aim of this article is to review the last twenty years of ‘machine listening’1 to sound and music, and to suggest a balanced approach to the human–machine relationship for the future. How might machine listening, and MIR2-based ideas of data storage, retrieval and presentation enhance both our embodied experience of the music and its more reflective study (analysis)? While the issues raised may be pertinent to almost any music, the focus will remain on electroacoustic music in its many forms, whether for interactive composition, performance or analytical endeavour. I suggest a model of listening with – that is, alongside – machines in such a way that our skills may be enhanced. What can we share with machines to mutual advantage?