I mean, if I understand it correctly, audio signals are additive so you don’t need to do fancy instrument changing. Simply add the sample of different instruments to the audio buffer and those can be played. If someone knows better about acoustic theory then please correct me.