One of the factors that must be considered in crossing over multiway speakers is the Woofer Highpass Offset. I discovered this recently reading Linkwitz.
But I was also thinking about it wrong, or not as Linkwitz does.
I was thinking what needs to be compensated for is the 180 degrees of phase lead at the lowest frequencies near the lowpass of the woofer.
That's a huge amount of delay!!! 32 feet at 16 Hz for example.
Alternatively, I had thought of 180 degrees remaining constant to the crossover frequency, where it becomes a mere 5.5 ms.
Linkwitz looks at the change in delay with frequency (it is not constant). For his example woofer, it drops to 23 degrees at 100 Hz, so we're down to about 1ms of delay (right in my typical adjustment range even not knowing about this source of delay).
So time alignment has to be done by audio measurement, not physical measurement or intuition. (I wouldn't recommend listening at first either. The "correct" value should be first be determined by valid measurements, and then fine tuned by ear.
I knew this was true, but thought the problem stemmed from different crossovers (it still may, if the crossovers are optimized for EQ). Instead, the fundamental delay problem comes from drivers. They have internal phase shifts.
Still, I wonder, is the crossover frequency necessarily where the optimization needs be done? If the woofer really were "leading" the tweeter, it seems that one might want to optimize the leading edge, the first contribution, the envelope delay to the leading edge of an impulse.
Contrarywise, the crossover is where the drivers have to be in phase, etc, for the crossover to work as desired, otherwise it will have various issues.
Now, it seems like the woofer can't actually lead the tweeter by as much as we've described in real time. This phase lead must be slightly bogus somehow. I'm thinking the phase "lead" is relative to the woofer's own high frequency response.
So what this means is that with any woofer, when set to optimize the crossover, there is going to be a leading low frequency pre-ring, representing the woofer's frequency differential phase.
I think I've seen that. If aligned to the low frequency phase, there may be a hole in the midbass.
I think I've seen that also.
But I was also thinking about it wrong, or not as Linkwitz does.
I was thinking what needs to be compensated for is the 180 degrees of phase lead at the lowest frequencies near the lowpass of the woofer.
That's a huge amount of delay!!! 32 feet at 16 Hz for example.
Alternatively, I had thought of 180 degrees remaining constant to the crossover frequency, where it becomes a mere 5.5 ms.
Linkwitz looks at the change in delay with frequency (it is not constant). For his example woofer, it drops to 23 degrees at 100 Hz, so we're down to about 1ms of delay (right in my typical adjustment range even not knowing about this source of delay).
So time alignment has to be done by audio measurement, not physical measurement or intuition. (I wouldn't recommend listening at first either. The "correct" value should be first be determined by valid measurements, and then fine tuned by ear.
I knew this was true, but thought the problem stemmed from different crossovers (it still may, if the crossovers are optimized for EQ). Instead, the fundamental delay problem comes from drivers. They have internal phase shifts.
Still, I wonder, is the crossover frequency necessarily where the optimization needs be done? If the woofer really were "leading" the tweeter, it seems that one might want to optimize the leading edge, the first contribution, the envelope delay to the leading edge of an impulse.
Contrarywise, the crossover is where the drivers have to be in phase, etc, for the crossover to work as desired, otherwise it will have various issues.
Now, it seems like the woofer can't actually lead the tweeter by as much as we've described in real time. This phase lead must be slightly bogus somehow. I'm thinking the phase "lead" is relative to the woofer's own high frequency response.
So what this means is that with any woofer, when set to optimize the crossover, there is going to be a leading low frequency pre-ring, representing the woofer's frequency differential phase.
I think I've seen that. If aligned to the low frequency phase, there may be a hole in the midbass.
I think I've seen that also.
No comments:
Post a Comment