EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 1 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing EE2F1 Speech & Audio Technology Martin Russell Electronic, Electrical & Computer Engineering School of Engineering The University of Birmingham

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 2 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Speech & Audio Technology  Lectures – Tuesdays, 10am LT1  Two laboratory sessions: –NG22, Thursday/Friday 2pm-5pm –Week 6 – introduction to SFS –Week 8 – speech synthesis –Check timetable for groups  Examination next semester  Revision lecture next semester  Follow-on course EE2F2 Music Technology by Dr T Collins

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 3 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Demonstrations  Time-domain and Frequency-domain representations of sounds –Time-domain: ‘waveform’ or ‘signal’ –Frequency domain: ‘spectrum’ –Time-frequency representation: ‘spectrogram’  Tools –Sound Forge XP 4.5 (Commercial software) –SFS (Speech Filing System) 4.28 for windows –SFS is free!

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 4 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Sine wave 880Hz

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 5 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Recorder ‘A’ (880Hz)

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 6 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Piano middle C, +8ve, +2x8ve

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 7 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Saxophone

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 8 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Voice plus background ‘drone’ Kate Rusby, Annan Water from “Hourglass” Pure records 1997

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 9 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Rock Music Bruce Springsteen, Born in the USA, Columbia 1984

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 10 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Electronic music Moby, Porcelain, Mute Records 1999

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 11 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Analogue-to-Digital Conversion  A/D conversion  or ‘Sampling’  or ‘Digitization’ PC Sound Card Amplifier D/A converter Disk

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 12 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Pulse Code Modulation (PCM)  How many quantization points?  How many samples per second (sample rate)? Quantization error Sample point Quantization point

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 13 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Nyquists Theorem  (Ignore quantization error for now)  Suppose a signal has no frequency components above B Hz  If that signal is sampled 2B times per second, then it can be faithfully reconstructed from the digital signal  B is called the Nyquist frequency  2B is called the Nyquist rate  Reconstructing the original signal from the digital signal is called Digital-to-Analogue (or D/A) conversion

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 14 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Aliasing  Aliasing happens when a signal is sampled at a rate less than the Nyquist rate  Detail is lost in the reconstruction of the signal  Spurious low-frequency components appear in the reconstructed signal  These are “aliases” of real components of the signal  Example…

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 15 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Aliasing Nyquist rate Less than Nyquist rate Reconstructed “alias”

EE2F1 Speech & Audio Technology Sept. 26, 2002 SLIDE 16 THE UNIVERSITY OF BIRMINGHAM ELECTRONIC, ELECTRICAL & COMPUTER ENGINEERING Digital Systems & Vision Processing Analogue-to-Digital Conversion  For a given sampling rate, an “anti-aliasing” filter removes all frequency values above the Nyquist frequency PC Sound Card Amp D/A converter Disk Anti- aliasing filter