Lecture 37 Blog, ECE438 Fall 2014, Prof. Boutin

Friday November 21, 2014 (Week 13) - See Course Outline.

Jump to Lecture 1, 2, 3 ,4 ,5 ,6 ,7 ,8 ,9 ,10 ,11 ,12 ,13 ,14 ,15 ,16 ,17 ,18 ,19 ,20 ,21 ,22 ,23 ,24 ,25 ,26 ,27 ,28 ,29 ,30 ,31 ,32 ,33 ,34 ,35 ,36 ,37 ,38 ,39 ,40 ,41 ,42 ,43 ,44


Today we defined formants and discussed how to obtain the formants from the DFT of a sampled, finite duration, voiced phoneme. Recall that a voiced phoneme is the convolution of a pulse-train and and the unit impulse response of the vocal tract (assumed fixed shape). Thus, the CT Fourier transform of an analog recording of a voiced phoneme is the product of the vocal tract frequency response and the Fourier transform of a pulse-train (a sequence of impulses multiplied by Fourier series coefficients). After discretizing the voiced phoneme, we saw that the DT filter corresponding to the vocal tract is an all pole filter. This is because, according to our sequence of tube model, the transfer function of the DT system representing the vocal tract is rational and has no zeros.

Action Items

  • Finish HW 10. It is due Monday.


Previous: Lecture 36 Next: Lecture 38


Back to ECE438 Fall 2014

Alumni Liaison

Ph.D. 2007, working on developing cool imaging technologies for digital cameras, camera phones, and video surveillance cameras.

Buyue Zhang