Abstract
A synthetic signal was formed by passing a burst of energy through an all-pole digital filter, and Gaussian noise was added to the output. (This is the model of speech production that best recursive fit (BRF) was designed to analyze.) Both linear predictive coding (LPC) and BRF were used to try to recover the filter parameters; at all significant levels of noise BRF is an order of magnitude more effective than LPC. A similar experiment was done on real speech; both LPC and BRF were done (pitch synchronously) on voiced speech, then noise was added to the speech and LPC and BRF were done again. Recovered pole frequencies and bandwidths vary as noise is added; the expected “error” is less for BRF than for LPC. BRF analysis and resynthesis of speech will be demonstrated by the playing of a short passage.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
