ok, i just checked the cx on my Infinity BETA 50's ... they're at 600hz ... so they're good?
but the JBL E series for example has a low pass at 300hz ... that's bad?
granted manufacturers cut corners somewhere to be able to build cheap and sell plenty, but i'm having a hard time believing the engineers at polk would let something like this slide? (given the severity of the problem - as per your conclusion)
The speakers you quote continue to highlight the problem of three ways, and the need for bandpass drivers with larger bandwidths.
Now take the infinity first, the first crossover is 600 Hz, a nice place for a passive crossover. The next crossover is 3.3 K Hz where the optimal point 4.8 K Hz or above. So the compromise was to use fourth order crossover slopes to limit the band pass gain and get as flat a midband response as possible. What was the downside? Well it is time smear, because the mid is a full cycle ahead of the woofers at 600 Hz and the tweeter a full cycle ahead of the mid a 3.3 KHz. However given the limitations of the bandwidths of the drivers they are working with it was in my view the best compromise to make. However under controlled testing this amount of time smear has been shown to be audible. A speaker with analog fourth order crossovers can't even reproduce the vaguest facsimile of a square wave.
Now lets take the JBL the crossover points are 500 Hz and 3K Hz where the optimal would have been 500 Hz and 4 KHz. Again the compromise was fourth order crossover slopes.
Now in my three way I posted with the Dynaudio D76 mid driver, I get forced into the same compromise, even though the spread is 400 Hz and 4 K Hz. The reason is the limits of the out of band responses of the divers involved.
However those speakers are pleasing and they are not my critical listening speakers.
In my main system, all crossover below 1 KHz are active electronic crossovers, because of these issues.
If people really want to return to integrated full range speakers with high spl and low F3, then we need powered speakers with active crossovers, and preferably digital crossovers at that. There is still a crying need for good divers with significantly wider pass bands.
If I were a manufacturer, I would lock up the marketing guys and devote the lions share of my resources to full range driver development. The first sacred cow out of the window, would be rigid speaker cones. I believe this wrong headed obsession is the root cause of why we don't have a good selection of wide pass band speakers.
Members who have the time to doodle should follow the single driver full ranger sites. It is for the reasons outlined in these threads, I'm still a full ranger at heart, and so should every driver developer be.
The Polk compromise, in my view was a poor choice. That solution has to have a null at crossover and almost certainly accounts for the less than full bodied sound people have eluded to.
You have to understand, with what we have currently, there are numerous design compromises in EVERY speaker, no matter what it costs. Great speakers are to a very large extent dependent on the designer's judgment to make the best sum of the compromises he selects.
