Talk:Nyquist sampling theorem

Wording
Shouldn't the sampling rate be greater than double the highest frequency? Or is that not the correct technical definition? --196.35.158.180 22:53, 8 November 2009 (UTC)

From what I understand it is suppose to be half the frequency and any greater frequencies create a digital artifact issue referred to as "aliasing".This is remedied with a filter, but in order for the filter to be effective it needs an attenuation slope because a "brick wall" filter is technically impossible to create under the circumstances.Hence why a conventional sampling rates like 44.1Khz is not an even 44Khz

So (as far as I understand it) the theory that the sampling rate needs to be only be twice the rate of the source is correct, but the technical implementation and utility in real-world audio application and circumstances need to be accommodated and compensated for.

If you decide to use this info please double check and use verifiable sources... I am not an electrical engineer or a telecommunications guru by any means :) --Taoist 23:46, 8 November 2009 (UTC)

---

Just a short remark regarding the 44.1 kHz sampling rate: the reason for this rather strange rate is due to historic technical restrictions. When the CD was developed hard disk recording was not feasible, so video systems have been used for this pupose. The sampling rate is a result of video parameters. See e.g. http://www1.cs.columbia.edu/~hgs/audio/44.1.html