Hi,
first, i think thewas' post #4 has a lot of great info....
How does a speaker resonance manifest itself in the (acoustic) phase response? Is a speaker in an anechoic chamber considered a minimum phase system for its entire range? (lets consider a system without an electrical crossover for now).
When there's an anomaly in frequency response, it will cause phase to wonk out too, if for no other reason than the math behind the measurement ties the two together.
If the frequency response anomaly is within a single driver's operating range, it may or may not be a good candidate for minimum phase EQ, which would correct both freq response (more properly termed magnitude) and phase response.
The best way I know to determine whether the anomaly is a good candidate for EQ, is correct it on-axis, and then see how well the correction holds up off-axis. If it hold up well, I consider the anomaly minimum phase and suitable for correction, whether the anomaly is broad or narrow. Key for me, is simply does the correction work everywhere.
Yes, speaker is predominantly minimum phase for its entire range if it is a true full-ranger.
Sane for the pass bands of individual drivers.
IIR crossovers and their frequency ranges that have two or more drivers summing together, bust up minimum phase.
Also, in a broader perspective, I like to think minimum phase pragmatically ends, whenever multiple sources contribute to a frequency......like diffractions etc.
I think multiple sources breaks up minimum phase even when drivers are completely "in-phase" with each other on-axis .
Because, multiple sources entails multiple physical locations, which means "in-phase" only occurs a particular axis/distance/ spot of focus.
Should people use minimum phase filters or linear phase filters when it comes to manipulating a speaker's anechoic response? considering both options are available.
Well, I don't think linear phase filters should ever be used to manipulate a speaker's anechoic response,
because I think the only valid speaker EQs are minimum phase where when you fix magnitude and you also fix phase.
AND even then, those EQ's need to hold up both on-axis, and reasonably well off-axis.
Now, if by filters you include crossovers too, then I think complementary linear phase crossovers are an incredible (mostly unrecognized) blessing in speaker design.
But they don't really manipulate anything, they just help get the acoustic design to work as intended. So I'm not sure if this is part of the question.