That would depend on what kinds of aberrations they were. Diffraction cannot be minimum phase because it is by its nature multi-path. Some horn resonances might be MP, but take heed my warning below.
This discussion always gets back to the meaning of MP in acoustics. We normally think of MP as being correctable with a MP filter, no matter what the aberration is. But in acoustics that is not enough because even if one is able to correct the response at one point that does NOT guarantee that it will be corrected at another point.
My take on this is that the concept of MP is not very applicable to 3D problems like it is for two dimensional ones like in mechanics and electronics. In fact no one dimensional filter could ever "fix" an arbitrary 3D acoustics problem.
This discussion always gets back to the meaning of MP in acoustics. We normally think of MP as being correctable with a MP filter, no matter what the aberration is. But in acoustics that is not enough because even if one is able to correct the response at one point that does NOT guarantee that it will be corrected at another point.
My take on this is that the concept of MP is not very applicable to 3D problems like it is for two dimensional ones like in mechanics and electronics. In fact no one dimensional filter could ever "fix" an arbitrary 3D acoustics problem.
When you say a compatible driver, are you extending this to mean you'd want to line up minor driver resonances to oppose minor horn resonances? I think I largely agree with your point of view... but it almost seems as if you are beginning with the 'audiophile' view on filters, and then contradicting it?
Nothing that esoteric, in fact, part of my whole loudspeaker design philosophy is to refrain from depending on mechanisms where alignment is critical because stuff shifts around. I tend to do all system alignments, from cabinet to horn to crossover generally overdamped and using design approaches that are tolerant of parameter shifts.
What I'm talking about is the basic stuff, the electro-mechanical properties of the driver, the front and rear chamber, and the flare profile, including the part internal to the compression driver and transition to the horn flare.
I almost put a "period" at the end of the sentence without the qualification of the driver, saying, "A horn that provides reasonably good acoustic loading will not have peaks its response curve." But then I added to that, "when driven with a compatible driver", simply to be a little more complete. I did this just to be precise, because I think it goes without saying, but then again, I can mismatch a horn and driver and get pretty bad response.
It can be something like throat exit angles, or it can be electro-mechanical specs, phase plug or chamber sizes. Electro-mechanical specs are pretty compatible between 1" compression drivers, with whatever differences swamped by the horn. Then again, those changes become a little more significant when the acoustic load is weak. But in general, I think the electro-mechanical specs of all 1" drivers are "in the same ballpark".
This is not always the case when selecting drivers for larger horns, midhorns and basshorns, for example. The driver, front and rear chambers and flare profile all form a system, and they must be compatible. I haven't seen electro-mechanical specs throwing 1" drivers off, but I have seen bad throat matches and even stuff like leaky rear chambers making the response poor. So the whole "compatible driver" thing was just a qualifier.
I've worked with some many good horns that provide super smooth response, and the trade-off going to constant directivity was always the ripple and other anomalies. To me, the improvements of recent years have been the movement towards constant directivity devices with improved smoothness. I want the response smoothness of an old-school exponential horn, and the coverage pattern of a constant-directivity horn, but without the resonances, diffraction slots and hard edges that make 'em sound so nasal, spitty and harsh.
Last edited:
Yep 1000% agreed, though I'm not as notch-averse as you- of course, I'm tuning my systems to my drivers, every time, where you need repeatability even with small batch variance.
My experience is that the variations in drivers is not that great. I don't find parameter variations effects on the crossover to be such a big issue. Hardly on the same level as getting the crossover right in the first place.
Depends on the XO, though, right? A high Q notch would be very sensitive to small shifts in cones etc, with a metal-coned system like the seas excel that were so popular a few years back. Of course, only a few systems actually managed to suppress that shrieking resonance, I think part of that was the driver parameters shifting dynamically and making the notch "miss".
Badman,
I would tend to agree with you that a narrow notch filter would be very sensitive to any shift and that is why you generally see these notch filters having a wider band area so that this doesn't become a problem in production. True conjugate notch filters are seldom seen in production consumer applications, and this includes most audiophile speakers.
I would tend to agree with you that a narrow notch filter would be very sensitive to any shift and that is why you generally see these notch filters having a wider band area so that this doesn't become a problem in production. True conjugate notch filters are seldom seen in production consumer applications, and this includes most audiophile speakers.
Yes, of course. But something is wrong if high-Q filters are required. This is where I can agree with Wayne - if you find yourself using high-Q resonant filters then something is wrong somewhere.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.