My thanks to those of you who have mentioned my speakers, sometimes even in the same breath as SoundLabs. For the record, my bipolar models indeed deliberately seek to emulate some of the radiation pattern characteristics of the pre-PX series SoundLabs (the PX series has a narrower pattern than I can reasonably replicate). As I told Roger West, creator of SoundLabs, my aspiration is to build the second-best speakers... well, within their general price ballpark.
Duke
(disclaimer: SoundLab dealer & speaker manufacturer) |
Just as disparaging general statements made about planars overlook some interesting exceptions, so too disparaging generalities about horns overlook interesting exceptions. Despite the title of the thread, I think it's well within the spirit of Renmeister's inquiry for horn advocates to reply with recommendations that are likely to have similar qualities to his present system but without the bass integration issue.
Duke
dealer/manufacturer |
Renmeister, there is a "free lunch" available when combining low-damping-factor specialty tube amplifiers and speakers designed to work well with such. The low damping factor in effect changes the woofer's electrical damping in a way that increases bass output relative to what you'd get with a solid state amp. If the speaker designer anticipated this, he has tuned the box so that instead of the lower damping factor giving you a bass hump, it gives you more extended low bass.
Aside from dynamics, one of the things a good horn speaker does well is radiation pattern control, and as a result the reverberant energy in the room usually has a similar spectral balance to the first-arrival sound. This is one of the ways in which a good horn system emulates the behavior of live instruments, and is generally not a characteristic of more conventional speakers (though as one might anticipate from my post just upthread, there are exceptions!). I believe that minimizing the spectral discrepancy between the direct and reverberant sound reduces listening fatigue, and can explain why if anyone is interested. |
Clio09, the listening fatigue thing goes back to how the ear/brain system processes sound.
When a sound first reaches the ears, a copy of it is stored in a short-term memory. Then for the next twenty milliseconds or so, all other incoming sounds are compared with the sound(s) in our short-term memory, to see if they are NEW sounds or REFLECTIONS, that is, repetitions of the original signal. If they are reflections, then they're largely ignored as far as directional cues go. This is called the "precedence effect" or "Haas effect", after the researcher who first reliably described it. During this interval, reflections still contribute to loudness and perceived timbre.
Now the way the ear/brain system determines whether an incoming sound is a new sound or a reflection is by "looking" at its spectrum. If the spectrum is the same (or close to) that of a sound that's in the short-term memory, then it's a reflection. If it's obvoiusly different, then it's a new sound.
So, what about a reflection that's quite a bit different from the original? Well that hasn't been thoroughly investigated yet, but I believe we would find a continuum along which a spectrally-different reflection goes from "easy for the ear/brain system to correctly classify as a reflection" to "too distorted to be reliably classified". And I think that as the spectrum of the reflection differs more and more from that of the first-arrival sound, the more processing the ear/brain system has to do to correctly classify the reflection. In computer terms, it's like a task that uses up a lot of CPU power... and so the CPU heats up (we get a head-ache, after about 15 or 20 minutes): Listening fatigue. This isn't its only cause, but I believe it is a fairly common one.
Now remember that during the time the ear/brain system is ignoring directional cues from reflections, they are still contributing to loudness and perceived timbre. So a speaker may measure "flat" on-axis, but not sound "flat" because of the timbre-skewing contribution of the reverberant energy.
* * * *
Roscoeiii, I don't know the inner workings (including the impedance curve) of enough other manufacturers' speakers to give a really good answer to your question. Coincident comes to mind, and I think that KCS and PiSpeakers and Omega Loudspeakers and Tonian Labs are likewise strong candidates. Yours truly uses a variable-length port system so the box tuning can be adjusted for different amplifier damping factors (and different low-frequency acoustic environments). |
The effect on a bass system of a low damping factor amplifier is the same as raising the electrical Q of the woofer. If the bass system is designed in anticipation of this higher electrical Q, then the result is proper bass response. If not, then "muddy bass" will likely result. So the fact that muddy bass is the result in some instances DOES NOT MEAN that such will be the case where the system is designed to work well with that type of amp.
Or to put it another way that may be more consistent with audiophile terminology, speaker/amplifier matching is important. And not just in the bass region - the interaction of the amplifier with the speaker's impedance curve has implications across the spectrum. But that's another topic for another thread.
Pryso, I did not mean to imply that a good horn system has a small sweet spot when set up properly. In fact, I believe that a good horn system is capable of having an exceptionally wide sweet spot if it's designed with that intention. |
Mapman wrote: "The primary benefit of horns is efficiency."
No doubt I'm in the minority here, but my primary reason for using horns (waveguides being a type of horn) is radiation pattern control. |
Ime THD does not correlate well with subjective preference. Imo THD is not the right yardstick to be measuring with; a horn or amp or whatever can have excellent THD numbers and still sound pretty bad, or have bad THD numbers and still sound excellent. I'm not horn-bashing here; I'm THD-bashing.
Better yardsticks have been proposed, but the industry has ignored them. |
"Two complaints have always been...'horn coloration', which I 'claim' to be able to hear that, (to me) characteristic 'horn sound'. That and the 'lack of dynamic' consistency between the horn and the bass 'drivers'.
"There's this discontinuity that I've always heard between those different drivers.
"So I'll ask--(Duke if he's out there), does this still exist?"
Hi Larry, well I think the colorations and discontinuities can be brought below the detection threshold for most people. On the coloration front, it starts with the design of the horn itself. My preference is for constant-directivity types that introduce as little diffraction as possible; this type of horn is often called a "waveguide". Its use calls for a fairly complex crossover so it doesn't appeal to many purists, but if the designer does his job well, neither you nor your amplifier would ever guess that the crossover is complex.
Addressing the dicontinuity issue requres a slightly different paradigm than what most people are used to. Instead of pairing up the best horn and best compression driver with the best woofer, we need to pair up the woofer, horn, and compression driver that work the best together. It's like Jim Thorpe, Muhammad Ali, Joe Montana, Tiger Woods, and Bjorn Borg are perhaps the best we've seen in their respective sports, but put 'em on a basketball court and the worst team in the NBA would eat them for lunch. The key is teamwork.
So getting back to horn speakers, we want to match up the dynamic capabilities of woofer and horn as closely as possible. That may mean using a prosound-type woofer whose thermal compression characteristics can keep with the compression driver. That may also mean not using the most uber-magnet hyper-efficient compression driver on the market.
The big thing I keep coming back to is the radiation pattern, and that's because it goes back to the most basic thing that our ears home in on: The frequency response. Most of the in-room sound we hear is reflected sound, so we can't afford to pretend like it doesn't matter if high fidelity is the goal. In particular, a radiation pattern discontinuity in the crossover region will let you know that there's a transition between drivers. The on-axis response can be smooth; the phase response can be smooth; and yet we hear the transition... in that case, look to the off-axis response! It matters because it's a far larger part of what we hear than is generally appreciated. Horns give us the opportunity to match up the pattern in the crossover region so that the off-axis response doesn't have a glitch there, and imho that should be taken advantage of.
I've had fellow speaker designers step into my room at an audio show and remark that they couldn't hear the crossover. That's either a high compliment about woofer-to-horn integration in the crossover region, or an admission that their ears are shot! I also have a customer who had one of my woofer + horn systems in a nearfield setup, so close that you could literally lean forward and touch the speakers. I expected to hear a vertical discontinuity when I sat down to listen, but with eyes closed I honestly couldn't. The apparent sound source was the center of the horn, and the crossover frequency was about 1.6 kHz. Now there may well be some instruments that would have appeared to come from lower, but in the course of listening to several songs I didn't hear it. This was unexpected, but indicates that, in some caes anyway, you don't need to be far back from a horn system in order for the drivers to integrate well.
Not that mine are the only systems that do this by any means, but mine are the ones I can speak from experience about. So take all of the foregoing with as many grains of salt as needed. This is all imho and ime and ymmv and etc.
Duke |
Macrojack, this is a cat that can be skinned more than one way.
At first glance, it looks like a crossover of 350 Hz between a fairly narrow-pattern horn and a direct-radiator woofer would not have a good radiation pattern matchup, and I think is probably the case (though the horn may be unloading somewhat down that low, resulting in some pattern widening). But it may not matter if the crossover is done right.
You see, below 500 Hz in most rooms, the ear is not very good at hearing radiation pattern discrepancies as long as the power response is good (this is my understanding of Earl Geddes on the subject). Bill Woods knows far more tricks of the trade than I do (he has many fine prosound designs to his credit, some of which show up in home audio systems because they're so darn good). So I'm sure he has this transition worked out very well.
Turning now to the realm of dynamic matching, Bill uses woofers that have excellent dynamic capabilities and so their thermal compression will be negligible at any SPL remotely likely in a home audio setting, thus matching the compression driver in that regard.
Duke |
"Duke's speaker is not an Horn .... Waveguide !!!!!!!"
A waveguide is a type of horn, characterized by constant directivity, minimum-wavefront-disturbance curvatures, and a lack of vanes or diffractive shapes or features to modify directivity. Some of my models use what I would definitely call a "waveguide", and others use a device that has many (but not all) of the characteristics of a pure waveguide, so I call it a "waveguide-style horn". Macrojack's conical horns, designed by Bill Woods, are probably what I'd call "waveguide-style horns". Others may use different terminology and/or draw their lines of demarcation in a different place. For example, Bill Woods called my waveguide a "conical horn".
Duke |
"Seriously, I've GOT to hear some horns, just to revisit...Duke, if you ever want to have a Beta Test Site...send me a pair...I'll write a Harry Pearson quality, prosaic piece and be honest to a fault."
Thanks for your offer, Larry. Shipping costs make it impractical... I'd probably be looking at ballpark 700 bucks round trip.
Duke |
In three weeks the Lone Star Audio Fest is happening in Dallas, Texas. Here's a link:
http://lonestaraudiofest.com/
What this has to do with horns is, historically a disproportionate number of rooms at LSAF exhibit horn designs. We also have direct radiators, electrostats, dynamic dipoles, and probably others that I can't think of right now. The show was founded by Wayne Parham of PiSpeakers, and past exhibitors include Classic Audio Reproductions, Bill Woods, and Earl Geddes.
Duke |
Thanks for letting me know, Learsfool; I don't know how it got garbled so I don't know how to keep it from happening again.
Suggest you google "Lone Star Audio Fest", or type in "www.lonestaraudiofest.com".
Duke |
"...there's nothing important enough about audio to make someone angry--it's a hobby."
Very well said, Larry.
I think if people met in the hallway at an audio show, instead of on this internet forum / virtual battlefield, they'd instantly become friends.
Actually, I had just that experience a couple of years ago... meeting someone at RMAF that I'd squabbled with (and frankly been rather rude towards). He posted earlier in this thread, and he's a helluva nice guy.
Duke |
I can discuss technical issues but am not very good at emotional exchanges, so I will bow out.
Duke |
