Horn Question

[joyspring]

Whitebroncoii you seem to be refusing to consider my point, that a highly asymmetric, transient waveform cannot be reconstructed by drivers connected in opposite polarity.

Yes, they can depending on the amount of phase shift imparted by the low-pass and high-pass sections of the crossover and the amount of physical displacement of both drivers. Adjusting one to suit the other is the goal.

Phase, polarity and time-alignment are all separate but interrelated parameters.

I have experimented with this phenomenon for years, with many Altec and earlier Lansing systems using stock second order filters. Connecting the drivers in intended (reverse) polarity yields smooth response through the crossover region, at the cost of muddled imaging and degraded transients. Reversing one of the drivers results in a hole in the response at the crossover frequency, but provides far superior soundstaging and transients, where the system becomes far more invisible, or "immediate" if you will.

A few variables here, the first of which you do not detail: the physical displacement of both drivers in relation to each other...

Also, the type of filter alignment (Butterworth, Bessel, Linkwitz-Riley, etc.) will have an effect on the response because of the amount of phase shift, the amount of attenuation at the crossover point and of course the aforementioned physical relationship between the two drivers; the response (assuming the drivers are flat at and beyond the crossover frequency) is shaped the the constructive and destructive interference that occurs between the two as a net result of all of the above.

This is especially true if the system polarity agrees with that seen by the microphone(s) at the time of the recording.

Personally I'm not convinced about the audibility of absolute phase on complex waveforms though I do keep track of it during classical recording. I thought I could discern a difference on woodwinds though that was through sighted (not blind) evaluation which is essentially useless.

BobR