COMD 5070 Final Exam Questions and Answers
Graded A+
4 features of the scientific method - ANSWER-Empirical: based on data
Deterministic: Obeys physical laws, things are not random
Predictive
Parsimonious: Use simplest explanation but no simpler
How many semitones in an octave? - ANSWER-12
What is the Nyquist frequency? - ANSWER-Half the sample rate
The highest frequency the recording is able to reproduce
High pass filter - ANSWER-Allows high frequencies through
Low pass filter - ANSWER-Allows low frequencies through
Band pass filter - ANSWER-Allows a band of frequencies through - holds back higher
and lower frequencies
Band reject filter - ANSWER-Holds back a band of frequencies - allows high and low
frequencies through
What is the Fourier theorem? - ANSWER-All periodic sounds are made of a
combination of sine waves. Amplitudes, phase angles and frequencies all vary
What does a Fourier transform create? - ANSWER-A spectrum from the time domain
waveform
What is the x-axis on a time-domain waveform? - ANSWER-Time
What is the y-axis on a time-domain waveform? - ANSWER-Amplitude
What is the x-axis on a frequency-domain line spectrum? - ANSWER-Frequency
What is the y-axis on a frequency-domain line spectrum? - ANSWER-Amplitude
What is the x-axis on a 3D spectrogram? - ANSWER-Time
What is the y-axis on a 3D spectrogram? - ANSWER-Frequency
What does darkness represent on a 3D spectrogram? - ANSWER-Intensity
What is a Fast Fourier Transform (FFT)? - ANSWER-Revealing of source.
, Shows harmonic energy and range of harmonics in the sound.
What is Linear Predictive Coding (LPC) - ANSWER-Revealing of filter.
Shows spectral envelope and where the resonant frequencies are.
What does a line spectrum show? - ANSWER-A snapshot in time
What does a spectrogram show? - ANSWER-Speech over time
What is perturbation? - ANSWER-Rapid and random, i.e. hoarseness
What is modulation? - ANSWER-Rhythmic change in F0 and amplitude, i.e. vocal
tremor
Tracheal puncture - ANSWER-Direct, accurate measure of subglottal pressure during
speech
Medical procedure
Esophageal pressure - ANSWER-Direct measure of subglottal pressure
Person swallows esophageal balloon
Not a very practical or common procedure
Estimating subglottal pressure - ANSWER-Time when the pressure in the mouth is the
same as below the larynx
Finding intraoral pressure during a particular time - say a series of 'papapapa' and make
your measurements from that
How do you calculate average air flow? - ANSWER-if 1 liter of air is used...
and phonation lasts for 5 seconds...
average flow = 1/5 liter per second
or 200 cc/second or 0.2 liters/second
flow at a given instant may be different
Ohm's law - ANSWER-E=IR
Voice, flow and pressure are all linearly related to one another
Estimating laryngeal resistance - ANSWER-Psub (cmH2O) divided by flow (L/s)
semitone standard deviation [how do the numbers reflect our perception of intonation in
speech?] - ANSWER-Comparing SD in Hz would be fine if the mean were the same for
all speakers
Converting to semitones allows easy comparison across speakers - semitones are
scaled appropriately so that they reflect proportional changes
using semitones corresponds more closely to our pitch perception - not linear
Source-Filter Model - ANSWER-Larynx serves as the source
Vocal tract is the filter
Graded A+
4 features of the scientific method - ANSWER-Empirical: based on data
Deterministic: Obeys physical laws, things are not random
Predictive
Parsimonious: Use simplest explanation but no simpler
How many semitones in an octave? - ANSWER-12
What is the Nyquist frequency? - ANSWER-Half the sample rate
The highest frequency the recording is able to reproduce
High pass filter - ANSWER-Allows high frequencies through
Low pass filter - ANSWER-Allows low frequencies through
Band pass filter - ANSWER-Allows a band of frequencies through - holds back higher
and lower frequencies
Band reject filter - ANSWER-Holds back a band of frequencies - allows high and low
frequencies through
What is the Fourier theorem? - ANSWER-All periodic sounds are made of a
combination of sine waves. Amplitudes, phase angles and frequencies all vary
What does a Fourier transform create? - ANSWER-A spectrum from the time domain
waveform
What is the x-axis on a time-domain waveform? - ANSWER-Time
What is the y-axis on a time-domain waveform? - ANSWER-Amplitude
What is the x-axis on a frequency-domain line spectrum? - ANSWER-Frequency
What is the y-axis on a frequency-domain line spectrum? - ANSWER-Amplitude
What is the x-axis on a 3D spectrogram? - ANSWER-Time
What is the y-axis on a 3D spectrogram? - ANSWER-Frequency
What does darkness represent on a 3D spectrogram? - ANSWER-Intensity
What is a Fast Fourier Transform (FFT)? - ANSWER-Revealing of source.
, Shows harmonic energy and range of harmonics in the sound.
What is Linear Predictive Coding (LPC) - ANSWER-Revealing of filter.
Shows spectral envelope and where the resonant frequencies are.
What does a line spectrum show? - ANSWER-A snapshot in time
What does a spectrogram show? - ANSWER-Speech over time
What is perturbation? - ANSWER-Rapid and random, i.e. hoarseness
What is modulation? - ANSWER-Rhythmic change in F0 and amplitude, i.e. vocal
tremor
Tracheal puncture - ANSWER-Direct, accurate measure of subglottal pressure during
speech
Medical procedure
Esophageal pressure - ANSWER-Direct measure of subglottal pressure
Person swallows esophageal balloon
Not a very practical or common procedure
Estimating subglottal pressure - ANSWER-Time when the pressure in the mouth is the
same as below the larynx
Finding intraoral pressure during a particular time - say a series of 'papapapa' and make
your measurements from that
How do you calculate average air flow? - ANSWER-if 1 liter of air is used...
and phonation lasts for 5 seconds...
average flow = 1/5 liter per second
or 200 cc/second or 0.2 liters/second
flow at a given instant may be different
Ohm's law - ANSWER-E=IR
Voice, flow and pressure are all linearly related to one another
Estimating laryngeal resistance - ANSWER-Psub (cmH2O) divided by flow (L/s)
semitone standard deviation [how do the numbers reflect our perception of intonation in
speech?] - ANSWER-Comparing SD in Hz would be fine if the mean were the same for
all speakers
Converting to semitones allows easy comparison across speakers - semitones are
scaled appropriately so that they reflect proportional changes
using semitones corresponds more closely to our pitch perception - not linear
Source-Filter Model - ANSWER-Larynx serves as the source
Vocal tract is the filter
