With the Nyquist theory and sampling, when sidebands are generated, is the generation of the sidebands refered to as aliasing, or is aliasing a whole different matter entirely?
Can anyone help?
Cheers
Vanessa
Wed Feb 19, 2003 5:03 pm
AUdIoCoUrSeS
Joined: 31 Oct 2002
Posts: 2014
Nyquist and slopes
Vanessa
Aliasing is a term used to describe unwanted frequencies within a digital audio system. Think of the word "alias" outside of digital audio and it conjures up the meaning of "extra" or "additional" - i.e. an alias name.
Alias frequencies will be audible (when the binary is recontructed) because the original sampling rate was not at least twice the highest frequency.
Imagine a hill.
To measure the gradient of a hill (the steepness) we need two points on the hill so we can determine the "slope".
Imagine audio again.
The same can be said of audio frequencies, which also have a slope. Two points are needed (hence twice the highest f) in order to determine the slope rate (frequency) of the audio.
If the system only reads one point on the slope of the audio waveform it will mis-measure the slope, and hence add frequencies which are not actually present (alias frequencies).
Alias freqencies sound like "tweating" and are generally not musically related.
