[sdiy] group delay (of filters) and listening

[Magnus Danielson]

From: mirwin at qouest.net
Subject: Re: [sdiy] group delay (of filters) and listening
Date: Tue, 05 Oct 2004 11:28:15 -0400
Message-ID: <4162BD8F.7566 at qouest.net>

> Magnus,

Mike,

> Interesting description.

Thanks!

> So the effect of this will be that certain ranges of envelope attack times
> (and shapes) could be become severely distorted while others are relatively
> unaffected?

More or less, yes. The linear theory assumes exponential attack/decay patterns,
such that the amplitude has the form of A(t)=A0*exp(sigma*t) where positive
sigma gives rising amplitude, negative sigma gives falling amplitude and sigma
being equal will be for constant amplitude. The sigma-value is really the
x-axis position in the s-plane I was talking about.

> Something that would not be visible if slow frequency sweeps with steady
> amplitude were used for testing (except for steady-state peaks and dips).

Exactly! It's all in the _linear_ theory.

I have been trying to say this over and over for many years now.

Also, this knowledge should give you a direct indication of what in a piece of
music you should be paying attention to since this is where I expect the
difference to be.

The total sensation of a transient is both in amplitude and phase since they
are really just different aspects of the transient responce or rather impulse
responce. The phase as such for a constant signal is really not important - at
all! The behaviour in time is important for transients, and guess what our
music is full of!

Cheers,
Magnus