Lament is an interactive cabaret song in which skeletal upper body motion capture of the performer drives musical outcomes. The work draws upon the my background both as a classically trained soprano and informal experience in vernacular and experimental styles of vocalizing, including singing both in rock bands and in free improvisation ensembles. During the composition of this work, a family member died, and in response, I began to shape the work as a lament, a kind of intimate documentation of the process of making sense of a loss.
The movement-musical interaction facilitated by the motion sensors is an exploration of shifting musical and rhythmic cycles which sometimes only suggest comprehensible patterns. The musical response to movement becomes a kind of choreography in the work, affording different movements, including tautness, wrist snaps, and back flexibility. The work is a structured improvisation, in which movement and voice determine much of the melodic and rhythmic content, with some fixed musical elements. In the second section of the work, the violin notes triggered by hand movements are generated in real-time using a machine learning algorithm trained on melodies that I composed. Pitch, rhythm, and timbre of these notes are driven in part by performer movement.