| Résumé |
With recent advances in score-following capabilities, it has become possible to envision new timing strategies, to realize previously impractical methods of coordination between a live performer and electronics. Our work centers on the challenge of synchronizing at the end of a musical phrase, where events and processes are timed not from an initial trigger, but occur according to a relative distance to- wards a predicted future attack. The key software component is Antescofo, a score-following tool which relies on a strong coupling, through a dynamic programming language, between a machine listening module and a reactive engine. This language allows hierarchical, concurrent, and heterogeneous computer processes to be organized over time, and for external events to be anticipated, with a runtime system that triggers electronic actions in response to variations in performance. Various programming strategies were implemented, honed and tested with a live performer. The musical material of these sketches focused on the idea of flexible canons, on interactions between a live instrument and a second voice generated from a buffered real-time recording. The different canonic strategies make use of Antescofo’s live tempo calculations to create dynamic tempo shifts, to force voices to converge, and to control precise canon effects. |
