What’s the deal with speakers with horns?

[Swerd]

I'm talking about those speakers that use huge woofers and compression tweeters mounted on large horns. These drivers, intended for auditorium PA systems or guitar amplifiers, can get very loud, but may not be as clean sounding as many Hi Fi drivers. Many of these designs suffer from narrow dispersion, and at worst, audible horn resonances. In the past year or two, I’ve noticed a new fixation on these so-called high sensitivity (HS) speakers (people more commonly call them high efficiency speakers, but that’s a misnomer). What’s going on? Is this the most recent fashion trend, or is there something of real benefit?

My first reaction to all this was surprise followed by disbelief. My limited experience with these HS speakers at a few audio shows has left me with a very poor impression of them. To be fair, I listened under poor conditions in hotel rooms at the Capital Audio Fest. If I recall, one such speaker sounded good, but I can’t remember its name, and all of the others sounded quite bad. They all (except one) suffered from a horn sound or resonance that made them sound like a megaphone (for the lack of a better word). It would be unfair to say all these speakers shared the same design, so I can only make the limited conclusion that I have not yet heard a HS speaker I like.

These speakers have all or some of these possibly related features:

• High sensitivity (HS)
• High volume
• Big PA/instrument amp woofers
• Compression tweeters in horns
• Narrow dispersion (aka controlled directivity)

So I recently took the time to read the Earl Geddes white papers:

• What is a Waveguide?
  http://www.gedlee.com/downloads/What is a Waveguide.pdf
• General philosophy
  http://www.gedlee.com/downloads/Philosophy.pdf
• Directivity in Loudspeaker Systems
  http://www.gedlee.com/downloads/directivity.pdf

I’ll focus on the last paper about controlled directivity. What does he say, and what does he not say? Work has been dreadfully disappointing lately, and we are under assault from the finance barbarians. So, I’m enjoying some escape.

Geddes presented a different and possibly useful way to graph a speaker’s frequency response (FR) data as the angle varies between the speaker and the microphone. He claims that these complex multi-color plots show much more information than the standard method where several FR curves taken at several different angles both on- and off-axis are shown on one plot. He may have a point, but these graphs are so different from standard FR graphs, that I had to struggle before I could at least imagine I could understand what he claims they show. I am not sufficiently used to looking at them so I could quickly learn something useful after quick inspection. At present, I wonder if Geddes may be the only person who actually is familiar with these plots. I found myself wishing he had presented both standard and Geddes-type plots of several widely available speakers. That might go a long way in establishing whether or not his method is useful.

Geddes tries to make the case that his controlled directivity speaker design sounds better than standard speaker designs. He claims his multi-color plots show that. He even goes so far as to suggest that his plots can function as a surrogate measurement for superior sound. But he never provided data from blind listening tests showing that controlled directivity speakers are actually preferred by a significant number of people. Without that kind of data, his graphs are pretty pictures that don’t tell us anything significant.

What's your take on this or anything else about using auditorium PA speakers in a home?

 

[panteragstk]

I actually had this question come up many times when I sold PA equipment. Kinda the same as using pro amps in the home, but with more obvious conclusions. I've also read many a thread on AVS about this exact topic.

What I've been able to gather from people making the request for having PA speakers for HT use is that they think that they will be able to handle the sudden peaks easier than a dome, ribbon, etc. Why they think that is related to the high sensitivity that you mentioned. The argument on AVS is that "movie theater speaker systems use "horns" so why shouldn't they work in the home?" Well, from what I know of that type of speaker it is designed to be listened to from a pretty good distance away from the speaker. I have heard more PA systems than I'd like to admit, and not one of those speakers would be "good enough" for my HT. The newer QSC K series is the closest thing I can think of that may work OK. Don't get me wrong, there are PLENTY of fantastic PA speakers, but you wouldn't be able to fit them in most homes, and if you could you'd be WAY to close to them IMHO.

 

[TLS Guy]

I actually had this question come up many times when I sold PA equipment. Kinda the same as using pro amps in the home, but with more obvious conclusions. I've also read many a thread on AVS about this exact topic.

What I've been able to gather from people making the request for having PA speakers for HT use is that they think that they will be able to handle the sudden peaks easier than a dome, ribbon, etc. Why they think that is related to the high sensitivity that you mentioned. The argument on AVS is that "movie theater speaker systems use "horns" so why shouldn't they work in the home?" Well, from what I know of that type of speaker it is designed to be listened to from a pretty good distance away from the speaker. I have heard more PA systems than I'd like to admit, and not one of those speakers would be "good enough" for my HT. The newer QSC K series is the closest thing I can think of that may work OK. Don't get me wrong, there are PLENTY of fantastic PA speakers, but you wouldn't be able to fit them in most homes, and if you could you'd be WAY to close to them IMHO.

For classical music in the home compression drivers and horns are an absolute disaster.

However some of the pop crowd love them, especially those that want to recreate the 70s sound.

In addition some HT enthusiasts want to create the cinema sound. You really can't do that without sectoral horns and compression drivers. Personally I usually dislike the sound in cinemas and prefer the much more natural balance and reproduction of speech I can obtain in my own theater.

One thing for sure, those designs minimize thermal compression and those systems minimize distortion at sudden peaks and high power levels.

I think part of the issue is what people are sensitive to. Because of the brief power busts of program most are not hearing the distortion in their systems and don't realize how often their systems are actually running out of gas, even when overall spl is modest.

High powered non horn speakers really need robust motor systems and much more power that they realize to keep everything relaxed.

If I had a big horn powered system here 300 watts would be more than enough, whereas with standard moving coil speakers I use 3 KW to keep everything relaxed and smooth under all conditions.

 

[ski2xblack]

It's all a continuation of the horn tradition to me. PWK's Eight Cardinal Rules harken back to when home audio was purely science based. I view folks like Earl Geddes, Wayne Parnam, and others as carrying the torch with their work, all adding their own findings/research/opinions to the history. Reading the back and fort discussions between Wayne and Earl in the forums makes for an interesting read, and touches on a lot of the very questions you have. Fast forward to the information age, and the easy access to powerful design tools, add a group of industrious DIY-ers, and presto, you have things like the SEOS designs, in a wide variety of flavors, in a largely collaborative effort that includes some big brains with lots of experience, using everything from cheap drivers to some pretty pricey ones. I'm sure at least one or two of those creations sound far better than "PA speakers." I think that while catering to a narrow niche, the current state of things is an interesting development, to me at least.

On the practical side, I would disagree that high sensitivity drivers are less clear than lower sensitivity drivers. If anything, they have a leg up by moving less (lower driver induced distortion), and having much greater dynamic capabilities (which is of utmost importance for realism). Narrow/controlled dispersion helps manage room interaction in a way that wide dispersion cannot. And DIY is fun and rewarding, so there's that aspect.

 

[Swerd]

I actually had this question come up many times when I sold PA equipment. Kinda the same as using pro amps in the home, but with more obvious conclusions. I've also read many a thread on AVS about this exact topic.

Thanks for replying. I was beginning to be afraid no one would respond, nothing but crickets and tumbleweed .

Reading some of that stuff on the AVS DIY forum was what started me to wonder about this. I thought better than to post this question there because so many people have swallowed that kool aid.

What I've been able to gather from people making the request for having PA speakers for HT use is that they think that they will be able to handle the sudden peaks easier than a dome, ribbon, etc. Why they think that is related to the high sensitivity that you mentioned. The argument on AVS is that "movie theater speaker systems use "horns" so why shouldn't they work in the home?" Well, from what I know of that type of speaker it is designed to be listened to from a pretty good distance away from the speaker. I have heard more PA systems than I'd like to admit, and not one of those speakers would be "good enough" for my HT. The newer QSC K series is the closest thing I can think of that may work OK. Don't get me wrong, there are PLENTY of fantastic PA speakers, but you wouldn't be able to fit them in most homes, and if you could you'd be WAY to close to them IMHO.

Exactly what I think. Home audio has the listeners roughly 10-15 feet from the speakers. In a large theater or auditorium, the listeners can easily be 100-150 feet away. Speakers that produce narrowly dispersed sound (aka beaming or controlled directivity – different terms but same phenomenon) will spread their sound much wider at that distance.

Do you have any idea what contributes to that "horn sound" I talked about? I recently heard GedLee Abbeys (the horn lovers, excuse me, the waveguide lovers speak highly of these speakers) and they had that characteristic unpleasant sound. I would not want to have them at any price.

 

[fmw]

Klipsch has certainly been successful at integrating horn tweeters and midrange drivers into home audio speakers. Go listen to a pair of Klipschorns and you'll see what I mean. JBL also did some good horn speakers many years ago. They still use horns in their PA speakers. While PA speakers generally aren't very high on the fidelity scale, many pro audio speakers are without question. Genelec, Neumann, Focal and others make recording monitors that are nothing short of outstanding. Many of the lesser monitors are also excellent in home environments. My main home theater has Epos speakers for the mains. They are conservative sounding British made speakers of high quality. My center and surrounds, however, are relatively inexpensive Tannoy recording monitors. They do an oustanding job for what they have to do and mix nicely with my Epos mains. If you give it some thought you will conclude that the center channel is mostly voice and recording monitors handle that about as well anything. The surrounds are mostly sound effects - everything from crickets to helicopters. Recording monitors do this pretty well also. Where you really need the best sound reproduction is the mains where the music is delivered. For me it makes perfect sense to use high quality speakers for the music and lesser speakers for the rest of the sound track.

 

[HifiSystems1]

If you want to hear a brilliantly designed horn speaker- check out the largest offering from Oceanway. Allen Sides does an incredible job with them and when I heard them at CES they were the only speaker at the show with lower distortion than the Legacy Whisper XD. The Oceanway was of course much more expensive and a larger footprint, but still a very impressive engineering feat!

 

[ski2xblack]

Remember how AJ had a touring pair of his Soundfield Monitors going from home to home? This thread makes me think that Wayne Parnam or the diy-soundgroup folks should do the same thing with a Three Pi or one of the SEOS jobs. Let their wares make the rounds so folks like Swerd can smell what they're stepping in.

 

[Swerd]

Because I'm the OP, I'm going to stick with the original question, and will politely decline all offers to check out the latest horn speaker offerings. I want to understand what is it about the horn speaker sound I so much dislike.

Do you have any idea what contributes to that "horn sound" I talked about? I recently heard GedLee Abbeys (the horn lovers, excuse me, the waveguide lovers speak highly of these speakers) and they had that characteristic unpleasant sound.

When I heard these GedLee Abbeys, I heard some music that contained a prominent clarinet passage. It sounded like a clarinet being played in a large public restroom with tile walls and floors. I made an effort to walk all about the room and I heard this unpleasant sound no matter where I was. (I played a clarinet in high school, and I know first hand what this sounds like .) I was trying my best not to violate the number 1 rule of audio – Never Say Anything Bad About Another Man's Speakers While He's Invited You To Listen To Them – but when he asked what I thought, I had to work hard coming up with something irrelevant and distracting. I talked about how good his amp (vacuum tubes) and speaker cables seemed . (This guy doesn't like Audioholics, so I believe he won't see this .) Fortunately, I did have to leave soon, so I didn't have to keep up the fiction that I liked them for long. It took me about 2 seconds to know they were ear fatigue waiting to happen.

The Abbeys, are large 2-way speakers with what looks like a PA type 12" woofer and a 1" (?) compression tweeter mounted in a round waveguide that looks the same diameter as the woofer. The web site shows these two frequency response graphs, one is the usual type (albeit stretched out horizontally), and the other is the Geddes-type multi-color plot.

The crossover is apparently between 600 and 700 Hz. The upper graph looks reasonable – nothing unusual stands out that might contribute to "that horn sound". The Geddes-type graph shows a prominent ridge between 400 and 500 Hz where the sound dispersal becomes noticeably less dispersed, but that would be below the crossover frequency. So I am not sure I see anything that might be a problem.

I once read that horn mounted tweeters can generate a resonance whose frequency is related to the perimeter of the horn, but I can't remember anything else about that. Does this idea ring a bell with anyone? Is this at all related to "that horn sound"?

 

[Steve81]

I once read that horn mounted tweeters can generate a resonance whose frequency is related to the perimeter of the horn, but I can't remember anything else about that. Does this idea ring a bell with anyone? Is this at all related to "that horn sound"?

There was some mention of that when the RF-7II measurements from a German audio magazine were posted over at AVS; IIRC that particular speaker had some severe ringing around 400Hz or so. I think I mentioned to you at the time that I'd suspect it'd be less of an issue with the lesser models of the Klipsch Reference lineup (ironically) given that the horn in those models is integral to the front baffle, though obviously I haven't tested that theory.

 

[Swerd]

There was some mention of that when the RF-7II measurements from a German audio magazine were posted over at AVS; IIRC that particular speaker had some severe ringing around 400Hz or so. I think I mentioned to you at the time that I'd suspect it'd be less of an issue with the lesser models of the Klipsch Reference lineup (ironically) given that the horn in those models is integral to the front baffle, though obviously I haven't tested that theory.

I would guess, if a waveguide rim rang like a bell, that should occur at a frequency where the tweeter was making some sound . In the Abbeys, that's below the crossover frequency.

I know I heard something unpleasant, but I can't explain it. Maybe my guessing is wrong.

It must have been those speaker cables .

 

[jinjuku]

Thanks for replying. I was beginning to be afraid no one would respond, nothing but crickets and tumbleweed .

Reading some of that stuff on the AVS DIY forum was what started me to wonder about this. I thought better than to post this question there because so many people have swallowed that kool aid.

Exactly what I think. Home audio has the listeners roughly 10-15 feet from the speakers. In a large theater or auditorium, the listeners can easily be 100-150 feet away. Speakers that produce narrowly dispersed sound (aka beaming or controlled directivity – different terms but same phenomenon) will spread their sound much wider at that distance.

Do you have any idea what contributes to that "horn sound" I talked about? I recently heard GedLee Abbeys (the horn lovers, excuse me, the waveguide lovers speak highly of these speakers) and they had that characteristic unpleasant sound. I would not want to have them at any price.

Some of these designs aren't horns though Swerd. There is a difference between horn loading and wave guides. If you are a primary listener to your HT or 2 channel trade some dispersion for controlled directivity by all means.

I don't think there is any kool-aid drinking. Just preference. Look Jeff Bagby has made many a speaker. I'll give him and his SEOS benefit of the doubt.

 

[Steve81]

I would guess, if a waveguide rim rang like a bell, that should occur at a frequency where the tweeter was making some sound . In the Abbeys, that's below the crossover frequency.

Well in the Klipsch's case, the resonance at 400Hz would have been over an octave below the XO (which I'd hope would be pretty steep). Here's the thread in question, and the comment that raised the issue FWIW: Klipsch RF7 II Measurements

I'm thinking that these resonances may be contributing causes to the honkiness and listening fatigue that many people have experienced with modern Klipsch home speakers. That resonance at 400 Hz happens to correspond precisely to the outer circumference of the horn, which must not be adequately braced and damped, allowing the horn to literally ring like a bell. And the 1200 Hz crossover frequency for the high-frequency driver means that its colorations, whether detectable in these measurements or not, will be present throughout the midrange.

I know I heard something unpleasant, but I can't explain it. Maybe my guessing is wrong.

It must have been those speaker cables .

Hard to say without more info, but if they didn't use high grade power c(h)ords, that'd be my guess

 

[Dennis Murphy]

I'm agnostic on the subject, as I try to be about any design approach. I remember listening to a classic JBL horn monitor speaker in the 70's that was quite smooth and neutral. And I liked the original gigantic KlipschHorn. I would agree with Swerd, however, that most horn-compression driver designs I've listened to are unacceptably colored. I know one wave-guide, compression driver designer claims that its speaker is virtually flat 30-40 degrees off axis, and that the sound will be correspondingly accurate if the listener is positioned to take advantage of that characteristic. I confirmed that the response was flat at that angle, but the on-axis response was all over the place. And the sound was similarly colored no matter where I listened. Obviously the ear processes sound at a variety of arrival times, and one sweet measuring spot isn't going to translate into sweet sound. Still, I've agreed to work with very expensive high-sensitivity drivers with wave guides for a private party and for SEOS. I certainly hope it will be possible to mate the dynamic capabilities of this approach with decently accurate musical reproduction. If all a speaker has is the ability to play very loudly on peaks, it's more of a curiosity to me than a serious sound reproducer.

 

[ski2xblack]

Swerd, did the Gedlee you heard have those goofy sponges? The open-cell foam plugs were Geddes' ad hoc answer to horn coloration if I remember correctly.

 

[Swerd]

Well in the Klipsch's case, the resonance at 400Hz would have been over an octave below the XO (which I'd hope would be pretty steep). Here's the thread in question, and the comment that raised the issue FWIW: Klipsch RF7 II Measurements

Interesting. That's seems like the conversation I remembered. Thanks.

If that is the reason behind "that horn sound", it's either heard by some and not others, or it has been difficult to dampen.

If this is true, I would guess you could hear the resonance while playing only the woofer and while the tweeter was disconnected.

Could a waveguide be designed that worked at guiding waves without enough symmetry to also act like a bell? Or is it time to get out the plasticine?

Better yet, avoid the bell altogether and use a flush mounted tweeter on a flat baffle.

Swerd, did the Gedlee you heard have those goofy sponges? The open-cell foam plugs were Geddes' ad hoc answer to horn coloration if I remember correctly.

Yes, it had sponge foam plugs in the waveguide like in the web site photo.

 

[Steve81]

Better yet, avoid the bell altogether and use a flush mounted tweeter on a flat baffle.

Heaven help you if DS-21 catches you talking like that

Personally I don't have any particular subjective criticisms on the KEF UniQ array which effectively turns the midrange driver into a waveguide, though obviously YMMV. Of course, my complaints about the Klipsch setup that preceded the KEF lineup I'm now using were more to do with frequency response than resonance, so I may not be all that attuned to the problem.

 

[VicTorious1]

Heaven help you if DS-21 catches you talking like that

Watch out for that midrange mushroom cloud.

 

[ski2xblack]

Could a waveguide be designed that worked at guiding waves without enough symmetry to also act like a bell? Or is it time to get out the plasticine?

The SEOS folks already got out the plasticine:

The primary product that has emerged is the SEOS waveguide (Super Elliptical Oblate Spheroid describes the profile of the waveguide's curvature). The design objective was to provide good horizontal coverage for near constant sound anywhere in the room (controlled directivity for the audio wonks) and vertical coverage that would minimize floor and ceiling bounce, while minimizing the problems associated with traditional 'horns' (internal reflections and such)
​

Will that work better than foam plugs? Hopefully Dennis will share his impressions, he's the only guy posting who has (will soon have?) experience with both, and whose input I trust.

​​

 

[Irvrobinson]

GranteedEV used to be a proponent of the GedLees. Too bad he doesn't hang around here any more to comment.