Sound and video
Sampling, quantisation, the Nyquist-Shannon theorem, aliasing and a sound's bitrate; compression; video as a sequence of images, key and predicted frames; codec and container.
- 26 cards
- 5 topics
- Free
The outline, topic by topic
Each topic is weighted the way the exam weighs it.
- Concepts
- Questions
- Sort and link
- The numbers
- The curves
A few cards from the deck
Sampling: what is it?
Show the answer
Measuring a signal's value at regular intervals; the number of measurements per second sets its fineness.
A CD samples at 44.1 kHz: 44,100 measurements per second. Between two measurements the original signal is no longer known.
A recorder samples at 20 kHz. How high can a signal's frequencies go for it to be digitised without aliasing?
- Below 10 kHz
- Below 20 kHz
- Below 40 kHz
- Below 5 kHz
Show the answer
Below 10 kHz
The sampling rate must exceed twice the highest frequency, so that frequency must stay below 20 / 2 = 10 kHz. 20 kHz forgets the factor of 2, 40 kHz applies it the wrong way; above 10 kHz, frequencies would fold into lower ones.
By the Nyquist-Shannon theorem, to rebuild a signal without aliasing the sampling rate must exceed … its highest frequency; since the ear reaches up to about 20 kHz, a CD uses ….
Show the answer
By the Nyquist-Shannon theorem, to rebuild a signal without aliasing the sampling rate must exceed twice its highest frequency; since the ear reaches up to about 20 kHz, a CD uses 44.1 kHz.
Two measurements per period are the minimum to know an oscillation exists. 44.1 kHz leaves a margin above 2 × 20 kHz for the anti-aliasing filter.