24 Dec
2012
24 Dec
'12
4:47 a.m.
On Sun, Dec 23, 2012 at 10:06 PM, Chris Bannister
<cbannister(a)slingshot.co.nz> wrote:
So artifacts (transient peaks) on the trailing or
leading edges of a
square wave is not a form of ringing?
In this case, yes. They're a fundamental aspect of any bandlimited
signal and you'll see them even when the filters are all FIR. You can
see also them on any scope-- so long as the scope has a greater
bandwidth than the squarewave you're inspecting (perfect squarewaves
require infinite frequencies, and so do not actually exist).
And, like I said, you can massage the frequency response of a filter
to hide the ripples and overshoots by spreading the Gibbs effect over
a wider spectral area, giving the illusion that they're not there when
you look at it on a scope. That's mostly a trick of presentation.
Your ear doesn't care (though your circuits might, if you're trying to
limit the analog signal range).
You could also have a filter that really is adding ringing, but that's
something different.
Monty