wrote in message
...
Joel Kolstad wrote:
(I can't tell you how many times I've seen people stating something like,
'The Nyquist theorem requires sampling at at least twice the highest
frequency present in the signal," when of course it says no such thing.)

What do you think it means?

It means that perfect reconstruction of a signal requires sampling at at least
twice the _bandwidth_ of the signal present to insure that no aliasing occurs.

Two important points he

1) It's the bandwidth of the signal that matters, not the highest frequency
present (this is kind of the analog version of the digitial guys' "it's the
edge rate that matters, not the clock speed"). This fact is frequently used
to great advantage in radio receivers (and plenty of other designs, I'm sure).
2) The assumption that aliasing is inherently detrimental is not always true.
I've seen designs where well-defined bandpass filters were stuck in front of
an ADC and the aliasing was used _to advantage_ to let the ADC sample at much
closer to 2x then one could have obtained with more traditional filter design.
(Although I'd admit that this seems to have been more common when ADCs were
slower and you had to use all the tricks you could to get performance out of
them.)

---Joel