We discovered that light would not be appropriate because it would require musicians to focus on another visual cue / learn a new language. 

Even though we were two weeks away from the project deadline, we prioritzed listening to our users and pivoted by (1) exploring haptic feedback as output instead of light and (2) focusing on tempo and cuing in as main forms of gestural input (as opposed to stylistic elements such as crescendo/decrescendo).

Fortunately, there are a few case studies that support this new direction such as the haptic baton developed for blind musicians and the Pulse metronome watch. Because of these existing technologies, we felt confident in switching from light to haptic feedback.