Abstract: A music synthesizer is constructed according to a modular scheme with plural, substantially interchangeable voice units. During operation, these voice units are used to simulate different instruments. The voice units operate under control of a master computer, and take waveform data from a common memory through a common digital data bus. The actions of each voice unit in simulating a note are controlled according to a plurality of control parameters. These control parameters are derived by interpolating between plots of each control parameter versus time for a weak actuation (soft note) and a strong actuation (hard note) condition. The synthesizer is arranged to simulate the effects caused by the interactions between closely spaced excitations of the same instrument such as closely spaced strikes upon a drumhead, by varying the qualities of the sound. The synthesizer may also serve as a mixer or as a multichannel signal processing device.