Feb 1
Synthesis
In his master's thesis(1), Youngmoo Edmund Kim lists the most common forms of vocal synthesis and analysis. Those that combine analysis and re-synthesis, such as Sinusoidal Voice Coding, bypass any symbolic representation (notes, pitch, rhythm, etc.) of the audio input, and therefore are inappropriate for my project. Re-synthesis has to be from the symbolic data received from Python. FOF synthesis seems to be one of the most straightforward approaches. There is an FOF opcode in Csound which I'm trying to build off of. Once I get OSC messages working in Csound, I'll see how the system's output sounds.
Paper
I'm collecting sources and free writing to have some material to work with.
