research : Generating an Audio Synthesis Virtual Machine from Models

The standard architectural pattern for any non-trivial audio synthesis application is to factor out the synthesis engine into a virtual machine. Every new audio synthesis language or system ends up reimplementing the virtual machine. The building blocks of an audio synthesis virtual machine can be factored out into models. These could then be used to generate virtual machines that are the target for a given audio synthesis language.