On Mon, Dec 31, 2012 at 3:17 PM, Neil <djdualcore@gmail.com> wrote:
On Mon, Dec 31, 2012 at 2:57 PM, Charles Henry <czhenry@gmail.com> wrote:
> At the ends of the frequency range, the group delay becomes significant and
> shifts some components of the signal more than others.

So, the wider the band, the less likely you are to hear the shifted
parts of the signal?

Yes, in general. 
In the case we're describing, the top end of the passband is outside the range of hearing.  The phase shifts might still be audible at the low end of the transducer's range.
However, phase shifts aren't much of a problem by themselves--they're all over the place in any typical multi-driver system and placement in the room matters just as much.  The loudspeaker crossover introduces the same effect.

Subwoofers are much more prone to phase shift problems.  You notice it a lot when the bass seems to lag behind the beat--and this results from not having a continuous frequency response that goes down to 0 Hz (the ideal case) or bad resonance characteristics.

Really clean sounds have a lot to do with the tightness of the temporal response--but there are of course tradeoffs in any kind of design choice like these.

Chuck