Linear Analysis and Synthesis

Microcomputer Systems I (ECE-3551)
Dr. Veton Kepuska
Florida Institute of Technology, Melbourne FL


Author: Jason Rom
Date: Fall 2012

The purpose of this experiment is to reconstruct the speech using linear analysis and synthesis.  Our speech is produced from the larynx and other resonant modes which are filtered by the vocal tract and lips to create the sound that we hear when we speak. The goal is to separate the excitation signal from the filter. Linear analysis involves breaking down the voice into different parameters:  pitch, voicing, gain, and alpha coefficients. Synthesis would put all this back together in order to mimic the speech input. 

Files Included

  • C Code for DM6437
  • Project Report and Presentation