"The Artist" wrote in message

"Arny Krueger" emitted :

What's the point in a real-time EQ which "takes a few seconds to
take hold"? I have never seen or heard of such a thing.

I know of several examples.

Name some.

Given that they are in very common software like
Winamp, I'm surprised you've never encountered this situation. I've
seen it in hardware digital equalizers as well.

I don't use Winamp.

Then do us all the favor of not commenting on software that you don't use.

All the plugs I've used offer near instantaneous response.

The equalizer in Winamp isn't a plug-in. Please see former comments about
not commenting about something you have no relevant experience with.

In fact, it's a pre-requisite that they do so (have instantaneous
response).

Please see former comments about not commenting about something you already
admitted that you have no relevant experience with.

If I wanted a delay line...

If you had actually worked with the Winamp equalizer, you'd know what I'm
talking about. The core of most common digital filters is a tapped delay
line. The rest of that kind of filter is a mixer. This implies that the
filter has delay. The delay may or may not be frequency-dependent. Of
course, analog filters can easily cause signal delay as well.

Many digital filters have feedback, all IIR filters do. IIR filters are
chosen because they usually take less resources to accomplish a given
outcome. If you change filter parameters, it takes a while for the signal
levels in the various feedback paths to stabilize because the paths have
delays in them. Until the signal levels stabilize, the amplitude and phase
characteristics of the filter are in a state of flux. In many cases this can
be heard. It's especially audible in filters that affect low frequencies,
while it's less likely to be heard in filters that affect only the highest
frequencies.

Some digital filters are FFT-based. They aren't explicitly based on tapped
delay lines. Obviously, any change in the parameters of a FFT filter is not
going to be effective until the next batch of samples is processed. If you
want a narrow filter at low frequencies, say for rejecting hum, the sample
size is going to be significant. You again have a delay before changes
become audible.

Of course people who know the difference between things like FFT, FIR and
IIR filters know all about stuff like this. It's obvious that people who
don't know about what I'm talking about don't really understand even the
most basic topics in digital filter design. Furthermore, they must either
have very limited listening experiences, or just have ears that are
relatively insensitive.