And this has no demonstrated audible relevance of which I am aware.
I am familar with much credible perceptual reseach regarding audility of phase shifts upon music program signals, and in different circumstances. I am not aware of any such conclusion occurring in these texts, less I missed it. If you have reference to this, I would appreciate it so I can read up.
And beyond the visual aspect on a scope screen, what is the relevance?
4th order crossovers have not been shown to be detrimental to audio quality, but in fact, usually the opposite. The steep rate affords drivers to act, usually, in more ideal frequency ranges suited to them to minimize distortion, and resonances, and maximize off axis linearity/power response(in the case of mid/treble).
-Chris
The first person to draw attention to the detrimental effects of time smear was Edward Long in 1976. AES paper 1131.
I can't find the reference, but I read a paper a while back that was blinded using adjustable digital time delay. Interestingly experienced and inexperienced listeners set the controls, whose function they were not aware of for minimal time delay.
I think part of the problem is that time delay and phase aberrations are so common we are much too Cavalier about it.
I don't think speakers with large time delays sound accurate. It is hard to put the effect in words, but you know when it's there and when it not. In the speakers in question, the speaker is about 7ft ahead of the woofer at the 150 Hz crossover, and the leading edge of the tweeter is 2.2ft ahead of the mid at 500 Hz.
It really goes right to the hart of transient response. Peter Walker and I had a lengthy discussion about this one day. As you know his electrostatics were and still are highly time coherent. Peter Walker was never one to flog useless tweaks, but he convinced me this is an important issue.
I think a speaker that converts a square wave into a sign wave as all speakers with fourth order crossovers do, have to have significant limitations to their accuracy.
This issue is at the heart of the full ranger debate. A good full ranger like the JW module is quite a revelation played at sensible volumes. Unfortunately most full rangers use paper cones and are ragged. The JW is not. It has no cone break up modes at all. It does unfortunately have a slight suspension resonance at 325Hz, and a second harmonic at 750 Hz. There is first order roll off starting around 6 KHz, but no nasty peaks. It is a good tool for experiment.
My former monitors used first order time aligned crossovers. Since almost all of the time I was the only one listening, the lobing problem of a speaker with multiple drivers and first order crossovers was not a big problem. I would not recommend the exercise though, as the development time was 10 years.
The fact is though it has sensitized me to phase and time problems in speakers.
I think it is important to minimize these problems. I agree the fourth order analog crossover is tempting as in many situations it makes it easier to achieve flat frequency response. The drivers are so called in phase, but one must not forget the phase shift is 360 degrees, so the drivers are separated by a wavelength at crossover. For high resolution systems that is not a good thing.
