[Martin Taylor 970121 17:15]

Martin Taylor 970121 15:30]

I said:

Let's suppose you do absolutely no simulated operations on your samples. You

only want to reproduce the original waveform. You _can_ do this _exactly_

using samples taken just above the Nyquist criterion rate. For a sine wave

of frequency F, you can do if you have more than (1/2F) samples/sec.

That's an obvious (dimensionally impossible) booboo. It should read "if

the intersample interval is less than 1/2F seconds."

I don't suppose the error confused anyone, but it's as well to correct it

anyway.

Sorry.

Martin