An embodiment of the invention creates a method and system for assigning notes to be played by a musical synthesizer to a predetermined number of instrument voices available to be sounded by said musical synthesizer, so that the musical synthesizer may emulate the sound of a live orchestra or other ensemble. The method includes the steps of building an array based on the number of notes to be played and the number of instrument voices available to play such notes, and allocating notes to the voices pursuant to algorithmic determination. As notes are released or newly played, all notes are dynamically reassigned to instrument voices so that, to the extent practicable, all channels play almost all the time. Additional methodology provides for correct assignment of notes across multiple different sections (or types) of instruments for purposes of real time orchestration.