Revision as of 12:15, 22 November 2019 by BJH (Talk | contribs)

Formant Analysis Using Wideband Spectographic Representation

By user:BJH (Brandon Henman)


Topic Background


Audio Files

Custom-recorded, 16kHz sampling rate

Me saying "a" :Media:Vowels_voiced_a.wav ("ay")

Me saying "e" :Media:Vowels_voiced_e.wav ("ee")

Me saying "i" :Media:Vowels_voiced_i.wav ("eye")

Me saying "o" :Media:Vowels_voiced_o.wav ("oh")

Me saying "u" :Media:Vowels_voiced_u.wav ("yoo")


Methodology

Time Domain Representation of Recorded Signals:

SpeechSignals.png


Matlab Code


Spectogram Results & Formant Analysis

Wideband Spectogram of A:

WbSpec a.png

Wideband Spectogram of E:

WbSpec e.png

Wideband Spectogram of I:

WbSpec i.png

Wideband Spectogram of O:

WbSpec o.png

Wideband Spectogram of U:

WbSpec u.png


Summary


Alumni Liaison

EISL lab graduate

Mu Qiao