Process/update a single voice and return it's output and output to be reverberated
source not stored for this graph (policy: none)
nothing calls this directly
no test coverage detected