jon-myers
commented
7 months ago How much gradation do we need, and at which temporal intervals.
The easiest thing would be to just assign a level (from a preset of, say, 6 choices, mirroring pp, p, mp, mf, f, ff of western staff notation. ) for each traj. But, it would be nicer to be able to have these things vary over time: voices getting louder or softer during a trajectory. This requires something like an automation track in a DAW. Worth discussing our needs and options before going too deep into this. Would be good to have a simple way of assigning things, and then more discrete control for those users who need / want to use it.
Should have ways of assigning dynamic levels to each pivot point (or perhaps arbitrary points, like an automation track in a daw). This could perhaps eventually be driven by captured MIR data, but at first could be controlled by user