What is it? T_T
What is Aliasing?-physics watch
- Thread Starter
- 24-10-2017 15:33
- 31-10-2017 00:17
Aliasing is a sampling problem.
This is a phenomenon which occurs when you simply do not obey the nyquist sampling frequency criterion.
According to Nyquist, in order to successfully record the information from a continuous signal (analogue to digital conversion), you must collect samples at a frequency which is double that of the main signal itself. The idea is for you to be able to sample a wave at least twice within one complete wavelength.
If you decide to do anything less than that, you will end up sampling the wave at non identical phase positions within consecutive wavelengths. What this means is that you would end up with recording a totally distorted and inconsistent signal copy of the main signal. This would seem a bit like some random noise except that the actual information is totally lost for good.
This can be solved by anti-aliasing, either by slowing down the main signal, or speeding up the sampler. For this reason, we use coupling capacitors.
- Study Helper
- 31-10-2017 09:38
When reconstructing a digitised data stream (D to A conversion), if the sampling rate applied to the original analogue signal is less than twice the maximum frequency of the sampled signal (Nyquist sampling rate), a true representation of the original will be impossible to reproduce.
A classic analogy is the strobe effect. When viewing a rotating wheel spokes under a strobe light, the wheel appears to rotate backwards but the wheel is in reality, moving forwards.
Below, the diagram shows what happens when the sampling rate (black dots) is less than the original analogue signal frequency (red sinusoid). The reconstructed analogue signal (blue sinusoid) is produced from the stored samples as an 'alias' and erroneously lower frequency. There is no way of recovering the original from the stored samples because not enough information is available.
Last edited by uberteknik; 31-10-2017 at 17:10.