End-of-quarter projects should combine CLM synthesis, algorithmic control, and possibly some performance analysis and / or real-time execution. The sonic target continues to be singing voice. Groups have been assigned to study several of the known synthesis techniques for vocal simulation. Presentations will take place during the last two weeks, focussing on musical examples and explaining advantages or limits of the different methods.

Use of the web for documentation and especially presentation is encouraged, but above all -- musical examples -- (which may not end up on the web). If you send Chris the address of your project web document, he'll link it here. Avoid spending extra time to get web documents looking fancy. The in-class presentations should be your target.

By way of example, the linear predictive coding project might encompass the following topics:

• historical and recent musical examples (if any)
• Mathews, et al. Bicycle Built for Two (Bell Labs, 60's -- also the soundtrack to 2001)
• Moorer, Lions Are Growing (CCRMA, 70's)
• Lansky, Dodge, Kendall (various, 80's)
• use in speech synthesis
• sound examples of speech coding for telecommunications
• early text-to-speech vs. text-to-speech
• explanation of the method
• analytic excitation-resonance model, what does it take for perfect reconstruction?
• filter structure
• common excitation methods
• methods for preserving the residual (error)
• CLM examples
• real-time controller
• analysis / resynthesis of a note
• analysis / resynthesis of a musical phrase with algorithmic control of LPC parameters
• discussion of advantages / disadvantages for captuing aspects of singing voice
• models formant transitions well
• efficient
• but, consonants can end up in the residual
• lower-level controls are non-intuitive (filter coefficients)
• any commercial (MIDI) use of the technique
• probably not
• possibilities beyond simple vocal simulation
• extended vocal techniques
• super-bass, time-warping, modulated filters, etc.
• applications in non-vocal sounds
• bassoon example
• cross-synthesis
• talking bassoon