At 10:03 PM 4/12/2004 -0400, Steven C. Barr wrote:
>----- Original Message -----
>From: "Dave Bradley" <[log in to unmask]>
> > Um.  I'm confused.  Since EQ is accomplished by the actual shifting of
> > phase, how could the phase shift be quite low and have it still be an
> > effective EQ?
>As I understand it, equalization (or any variance in signal levels produced
>selective frequency filtering) does not specifically and intentionally
>accomplish its goal via phase shifting...rather, the fact that this
>is accomplished through reactive elements in the circuitry (inductance and
>capacitance) inevitably produces phase shift as a side effect.
>This brings up an interesting point, which would necessitate more knowledge
>theory than I posess. If equalization is accomplished through digital
>of the produced wave rather than through the introduction of reactive
>is there any "phase shift" of the resulting digital signal?
>Steven C. Barr

It is possible to shift phase without changing amplitude (filtering) or to
change amplitude without shifting phase. Simple circuits pair the two; more
complexity is needed to do one without the other. For example:

Between signal and ground drop a series capacitor followed by a resistor.
Drop another with the matching resistor in series with the matching
capacitor. Now pick up the signal across the points where the two series
elements meet. With 'infinite' load, the phase shifts 180 degrees with no
amplitude change from high frequency to DC.

Needless to say, that can be done even more flexibly in digital, but I'm
not sure how various tools (e.g., FFT) fit into the picture. In Adobe
Audition, if you use the "scientific filters" you can see the phase shift
versus amplitude plots for the various options offered.

[log in to unmask]