home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!munnari.oz.au!ariel!ucsvc.ucs.unimelb.edu.au!lugb!news
- From: MATGBB@LURE.LATROBE.EDU.AU (BYRNES,Graham)
- Newsgroups: rec.audio
- Subject: Re: DAT/CD Sampling Frequency: who cares?
- Message-ID: <1992Jul27.063553.5620@lugb.latrobe.edu.au>
- Date: 27 Jul 92 06:35:53 GMT
- References: <1992Jul24.091023.18052@elroy.jpl.nasa.gov> <1992Jul27.033213.25987@trl.oz.au>
- Sender: news@lugb.latrobe.edu.au (USENET News System)
- Organization: La Trobe University
- Lines: 30
- In-Reply-To: tpearson@titan.trl.OZ.AU's message of 27 Jul 92 03:32:13 GMT
- X-News-Reader: VMS NEWS 1.22
-
-
- In <1992Jul27.033213.25987@trl.oz.au> tpearson@titan.trl.OZ.AU writes:
-
- > chu@hanauma.jpl.nasa.gov (Eugene Chu) writes:
- > > In article <5994@otc.otca.oz> brendan@otc.otca.oz (Brendan Jones) writes:
- > > >in article <Jul08.171025.23033@yuma.ACNS.ColoState.EDU>, cc433336@LANCE.ColoState.Edu (Constantinos A. Caroutas) says:
- > > >> Theory says that if you can sample at least twice per cycle, you can
- > > >> recreate the signal exactly.
- > > >
- > > >No no no. Not twice per cycle. At least twice the highest *frequency*
- > >
- > > I think the sampling frequency must be twice the BANDWIDTH of the signal
- > > you're trying to sample. Of course, Nyquist never said it would be easy.
- > >
- > >
- > Stop the argument. These statements are all the same for a sinewave,
- > and they are all obviously trying to infer the same thing.
- >
- No, they aren't. The bandwidth of a sinewave is actually zero. If you KNOW
- it is a sine wave of given frequency, you only need to calculate
- a) the phase and b) the amplitude. So you need precisely TWO samples. Ever.
- If you don't know the frequency, you need a third.
- The bandwidth is the difference between the lowest and highest frequencies
- present. If there is no low freq component, the signal must precisely
- repeat after a short interval, so you can spread your samples over its
- repeats. For music (or anything with finite length), there are components down
- to zero. So the bandwidth=max freq present.
-
- But I agree the argument should stop.
- Graham B
-