COMD 5070 Exam 3: Questions With Easy-to-
Understand Solutions
4 features of the scientific method: Correct Answer - empirical,
deterministic, predictive, parsimonious
Octave Correct Answer - a doubling or halving of a frequency
How many semitones in an octave? Correct Answer - 12. each
semitone is a non-linear step in terms of # of Hz from one semitone
to the next. However, they all sound equal to us.
Sampling rate Correct Answer - The number of snapshots of
the sound in time. ( in Hz) 10 samples is a 10 Hz recording.
Generally, the higher the sampling rate, the higher the fidelity of the
recording. Graphic representation of the join the dots graph, the
more dots, the better the image, the fewer, the cruder. A higher
sampling rate will take up more space and require higher processing
power.
Nyquist frequency Correct Answer - half the sample rate. You
determine the highest frequency you want to include in your
recording and double it, this will ensure that the playback will go at
least as high as the original signal. Nyquist determines the highest
frequency you can reproduce in playback. CD sampling rate is
44,100 and it stores signals up to 22,050 Hz.
Aliasing Correct Answer - When sampling too slowly, you will
inaccurately record the original signal. High, original frequencies
will be improperly recorded as lower frequencies. Think of the sine
wave, where only certain dots were captured, so the original
waveform is "smoothed out" or aliased.
,How to prevent aliasing Correct Answer - This can be
prevented with a low pass filter at the nyquist frequency, to keep
out the high frequencies that might be misrepresented.
High pass filter Correct Answer - allows high frequencies and
attenuates low frequencies
Low pass filter Correct Answer - allows low frequencies and
attenuates high frequencies
Band Pass filter Correct Answer - Removes high and low
frequencies but lets through the frequencies in the middle
Band reject filter Correct Answer - Allows all frequencies
except for a specific band in the middle.
Fourier transform Correct Answer - a way to analyze complex
sounds by separating them into their components
How to do a Fourier transform Correct Answer - You take a
time domain wave form ( microphone signal - time on x and
amplitude on Y) you apply the fourier transform to it and it gives
you a frequency domain wave form ( line spectrum) which
represents a slice in time that has frequency on x and amplitude on
y. It's like parsing light through a spectrum which shows all of the
colors. So you can now hear all of the individual sounds that make
up the harmonic series - this is the spectrum / frequency/domain
display. Similar to a list of ingredients of a cake.
time-domain display Correct Answer - has time on the X axis
and Amplitude on the Y axis and shows a waveform that represents
sound directly. It shows air pressure changes over time
, Frequency-domain display Correct Answer - a line spectrum
that shows the frequency components of a periodic sound(at 1
particular point in time). Frequency is on the X axis and amplitude is
on the Y.
3D spectrogram features [what are the three dimensions?] Correct
Answer - X axis =time Y axis =frequency darkness =intensity
3D spectogram. what does it show? Correct Answer - it's a
hybrid of time and frequency domain. Showing how the strength in
each range of frequencies is changing over time.. Dark lines
correspond to the higher peaks from the spectrum. White is minimal
energy.
Y axis on a spectrogram is limited to Correct Answer - the
Nyquist frequency - because it is made from a digital recording
Wide band spectrogram (ex 300 Hz) Correct Answer - gives
clear time detail, poor frequency resolution - shows glottal pulses in
vertical striations.
Narrow band spectrogram Correct Answer - gives clear
frequency detail but poor time detail - shows harmonic details in
horizontal bands.
noise - on a spectrum Correct Answer - lots of vertical lines
representing all frequencies at equal amplitudes in random phase.
Spectral envelope Correct Answer - describes display that only
includes dots for the tops of the lines, not all of the vertical lines for
each sine wave
FFT spectrum Correct Answer - Fast fourier transform - clearly
shows range of harmonics that are present in a complex sound.
Understand Solutions
4 features of the scientific method: Correct Answer - empirical,
deterministic, predictive, parsimonious
Octave Correct Answer - a doubling or halving of a frequency
How many semitones in an octave? Correct Answer - 12. each
semitone is a non-linear step in terms of # of Hz from one semitone
to the next. However, they all sound equal to us.
Sampling rate Correct Answer - The number of snapshots of
the sound in time. ( in Hz) 10 samples is a 10 Hz recording.
Generally, the higher the sampling rate, the higher the fidelity of the
recording. Graphic representation of the join the dots graph, the
more dots, the better the image, the fewer, the cruder. A higher
sampling rate will take up more space and require higher processing
power.
Nyquist frequency Correct Answer - half the sample rate. You
determine the highest frequency you want to include in your
recording and double it, this will ensure that the playback will go at
least as high as the original signal. Nyquist determines the highest
frequency you can reproduce in playback. CD sampling rate is
44,100 and it stores signals up to 22,050 Hz.
Aliasing Correct Answer - When sampling too slowly, you will
inaccurately record the original signal. High, original frequencies
will be improperly recorded as lower frequencies. Think of the sine
wave, where only certain dots were captured, so the original
waveform is "smoothed out" or aliased.
,How to prevent aliasing Correct Answer - This can be
prevented with a low pass filter at the nyquist frequency, to keep
out the high frequencies that might be misrepresented.
High pass filter Correct Answer - allows high frequencies and
attenuates low frequencies
Low pass filter Correct Answer - allows low frequencies and
attenuates high frequencies
Band Pass filter Correct Answer - Removes high and low
frequencies but lets through the frequencies in the middle
Band reject filter Correct Answer - Allows all frequencies
except for a specific band in the middle.
Fourier transform Correct Answer - a way to analyze complex
sounds by separating them into their components
How to do a Fourier transform Correct Answer - You take a
time domain wave form ( microphone signal - time on x and
amplitude on Y) you apply the fourier transform to it and it gives
you a frequency domain wave form ( line spectrum) which
represents a slice in time that has frequency on x and amplitude on
y. It's like parsing light through a spectrum which shows all of the
colors. So you can now hear all of the individual sounds that make
up the harmonic series - this is the spectrum / frequency/domain
display. Similar to a list of ingredients of a cake.
time-domain display Correct Answer - has time on the X axis
and Amplitude on the Y axis and shows a waveform that represents
sound directly. It shows air pressure changes over time
, Frequency-domain display Correct Answer - a line spectrum
that shows the frequency components of a periodic sound(at 1
particular point in time). Frequency is on the X axis and amplitude is
on the Y.
3D spectrogram features [what are the three dimensions?] Correct
Answer - X axis =time Y axis =frequency darkness =intensity
3D spectogram. what does it show? Correct Answer - it's a
hybrid of time and frequency domain. Showing how the strength in
each range of frequencies is changing over time.. Dark lines
correspond to the higher peaks from the spectrum. White is minimal
energy.
Y axis on a spectrogram is limited to Correct Answer - the
Nyquist frequency - because it is made from a digital recording
Wide band spectrogram (ex 300 Hz) Correct Answer - gives
clear time detail, poor frequency resolution - shows glottal pulses in
vertical striations.
Narrow band spectrogram Correct Answer - gives clear
frequency detail but poor time detail - shows harmonic details in
horizontal bands.
noise - on a spectrum Correct Answer - lots of vertical lines
representing all frequencies at equal amplitudes in random phase.
Spectral envelope Correct Answer - describes display that only
includes dots for the tops of the lines, not all of the vertical lines for
each sine wave
FFT spectrum Correct Answer - Fast fourier transform - clearly
shows range of harmonics that are present in a complex sound.
