The final test will happen during class on Thursday (12/5). Here is a brief review.

Sampling Rate

- What audio property is sampled?

- How often do you need to sample to get an accurate recording?
- What happens if you don’t take measurements fast enough? (If you sample rate is not fast enough, or high enough.)
- What is the name of the frequency that is the limit of accurate recording?
- Standard CD Sampling Rate?

Bit Resolution

- Bit resolution affects the recording accuracy of what audio property?
- How much of an increase in accuracy do you get for each additional bit?
- in number of values?
- as dB added to Signal to Noise ratio?

- If an the audio property is not recorded accurately due to high enough bit resolution, what do we call the process?
- What do we hear when the audio property is not recorded properly?

Recording Chain for Analog Recording

Recording Chain for Digital Recording

(all of the above can be referenced in the link)

Domains and Domain Representations

- What parameter is missing from a time domain representation of an audio signal?
- What parameter is missing from a frequency domain representation of an audio signal?
- Granular synthesis is a process that takes place in what domain?
- Phase Vocoding and Convolution are processes that take place in what domain?

Identify Some Granular Synthesis Parameters

Granular Synthesis

- How does granular synthesis expand or compress the time duration of a signal?

Fourier Transform and Spectral Processing

- What is Fourier’s Theorem?
- The Fourier Transform does what?
- The inverse Fourier Transform does what?
- With the FFT, the size of the FFT is measured in what unit?
- The FFT size determines what analysis parameter?
- What is the relationship between frequency resolution and time resolution?
- How does Phase Vocoding expand or compress time duration?
- What are some musical uses of Convolution?

**Answers**

Sampling Rate

- Amplitude
- Twice per wave cycle, to get one positive and negative measurement.
- Aliasing
- Nyquist Frequency
- 44,100 Hz, or 44.1 kHz

Bit Resolution

- Amplitude
- increase
- double the number of values
- 6 dB

- Quantization
- Noise

Recording Chains: see link

Domains and Domain Representation

- frequency
- time
- time
- frequency

Granular Synthesis Parameters

- grain duration
- grain index/position in audio file
- panning
- grain rate
- playback ratio
- density/number of streams/number of overlaps

Granular Synthesis

- grains can be repeated to expand the time duration, or skipped to compress the time duration

Fourier Transform and Spectral Processing

- Any complex signal can be represented as the sum of some number of sine waves of specific frequency and amplitude.
- The Fourier Transform converts an audio signal from the time domain to the frequency domain.
- The inverse FT converts from the frequency domain to the time domain.
- Samples
- Number of Frequency bands analyzed
- Inverse
- Since the FFT essentially takes a static picture of the sound, you expand/compress with Phase Vocoding by varying the playback rate of analysis frames on resynthesis (the iFFT). The process is similar to changing the speed of film or video.
- Reverb, filtering, cross synthesis

## Leave a Reply