Why not horns?

Well, you'll likely have a one-note bass, but in a room like that, that issue is hard to overcome regardless of the technology.

I'm a big fan of ESLs and ribbons, but the horns I have in my own system are faster and more revealing, and less colored. I have heard plenty of horns that honk and do weird things- the throat and mouth design is everything in a horn, any errors and the distortion skyrockets. But if it is correct there is nothing out there that can touch them. I don't have bass horns, my system is a bass reflex/horn hybrid, about 98 db 1 watt/1 meter. The diaphragms in the horns are beryllium domes with a kapton surround with no breakups in the audio pass band, and the magnet structures are custom-built field coils.

IMO/IME field coil is the only technology out there that can be as fast and as revealing as ESLs, although I am very impressed by the High Emotion Audio tweeter, which is an interesting (and patented) cross between a horn and a ribbon, very efficient and very fast (and nearly omni-directional) without coloration. It might be the only exception to the superiority of ESLs and field coils.

I don't blame anyone for not liking horns, but to those who doubt based on their listening experiences, I have this to say: you really need to hear the right one, set up properly. Many horn drivers are highly reactive and don't work right if the amplifier used has a low output impedance due to large amounts of feedback! Just because a horn is expensive, don't trust that it is also designed right too! There are quite a few bloopers out there IMO.

Shadorne, Classic Audio Loudspeakers makes a horn that crosses over at 250Hz, owing to the kapton surrounds on the beryllium-domed drivers. The tweeter rolls in at 12KHz.

Only a few years ago, he used a TAD maple-machined horn, but it turned out to have an artifact (even with the TAD drivers) due to an error in the throat design. He had Bruce Edgar design a new horn, and no more artifact. The new horn is also a maple machined design.

This is what I mean by proper setup- a properly-designed horn with a a decent driver. Many of the horn systems I have heard lack these two elements, and so while being loud, are not particularly pleasant nor are they accurate. Once this proper setup is achieved, actually setting up the speaker in a room is quite easy, although I have to say that some of the larger rooms I've used in the last few years at shows have been problematic due to reverb in the room. But that is not really a speaker problem- that's just a room treatment issue.

The main issue that I have run into with horns is that they do poorly in a near-field situation. You have to have about 8-10 feet to make them work, so that you get proper correct blend between drivers. If you are that close to The Classic Audio speakers, you have to be sitting down or the tweeter output is missing. This works OK for me at home, despite a slightly smaller than average room, as my listening chair is at about 11 feet. Further back, in the dining room, I can be standing or seated and it makes no difference.

John Wolfe does wind his own field coils.

Eldartford, and Appollo66, I remember back in the early 90s Harry Pearson said about horns: "they aren't high end". He turned out to be wrong about that, even back then. Since then of course horns have continued to improve.

I enjoy mine not simply for volume. They are faster, more resolving and more cohesive than the best magnetic and ES planars I have heard. The same comparison holds true against regular cone systems; price is really no object either. I enjoy a good stereo, and I try not to be so snooty that I will not allow a good sounding system to be discredited just because the owner did not do it the way I would have- I try to keep an open mind.

So when I say I prefer the horns I do, its over mbls, Wilsons, Sound Labs, Quads, Magnaplanars, Appogees (and their current clones, which are quite good), Avalons, Dalis and many other excellent speakers too numerous to mention. Further, I don't expect horns to always be on top; all technologies improve and one has to be ready to pick up and move if you want to stay on top :)

Right now an excellent example of technology on the move is the High Emotion Audio loudspeaker, the same people that were behind the Pipedreams from 10-15 years ago. They have a new tweeter technology that, if described in a nutshell, is a cross between a planar and a horn, literally its a planar with the shape of a horn and so its about 112db 1 watt/1 meter. Just like the best horns, its also very very fast and super detailed- maybe the fastest I have ever heard (although not bright at all). Although it is not a traditional horn, at the same time it should not be excluded from this conversation, although the actual applications I have heard it in have only been about 92-93 db efficient monitors.

So in a nutshell, the reason one might choose a horn is that the right ones most definitely do work quite well, regardless of one's prior preconceptions. I know this because this is where I started from too- thinking that horns don't work for high end applications. They won me over, kicking and screaming.

Phase can be adjusted to a fair degree in the crossover as well.

Mapman, the horns I listen to, as far as the soundstage is concerned, seem entirely recording-dependent as far was whether they are upfront or laid back. IOW they are neutral in this department. Some recordings will have sounds that seem inside the speaker. Others, the speaker seems to not exist as far as the music is concerned.

Herman, I leave the design of speakers to someone else to prevent headaches. I was just pointing out that the crossover can be used to correct time alignment (sometimes this results in the need to reverse the phase of certain drivers as a result), and in fact a crossover will introduce some time alignment issues that might have to be otherwise addressed. I think its a bad idea to rely on that exclusively- doing some physical alignment is important too.

Unsound, I appreciate your last comment. It is where I was for many years, and to this day (and I expect for years to come) I still run into horn designs that are guilty of everything you mention in your first paragraph. When horns suck, they really suck. But like any technology, its all in the execution, and when you encounter good horns, they will lack most of the objections you have except size and maybe what it means to be ugly.

A special point with regards to

so sensitive that they amplify every minutia of noise and distortion

Really, every speaker must play every minutia of noise and distortion to be a good speaker, but the noise/distortion floor of electronics is definitely an issue with horns. When we first encountered customers with horns, I think we failed pretty badly on this point because our reference speakers at the time were 89 db. This forced us to really figure out how to get our noise and low level distortion floors down with the end result that we now build a much better product. That revealing nature is thus a boon to progress, but a bane if you have problems in the electronics.

One thing that Duke did not cover is the drivers in use in the horns. Many speakers have breakups which contribute to distortion; if a horn driver exhibits this the result can be really bad. I know of at least one driver that has no breakups in the passband (its a true beryllium dome) and is helped by a kapton suspension that prevents artifact at the edges of the passband (250Hz-12KHz). This driver is very fast, very detailed and very smooth.

The other area that can be a problem is the interface of the driver to the horn- the throat area. The 2" TAD driver and 500Hz machined maple horn were an excellent example of this problem- right at the crossover/passband limit there was a peak followed by a dip before it smoothed out and it was not pleasant.

BTW this was in the Classic Audio Loudspeakers up until about 2 years ago when a new horn and driver (field coil) was introduced that fixed that problem (Valin gave the new driver on the old TAD horn Best Sound at Show and a few months later when the new horn was added declared that the new horn/driver combo was 'the biggest improvement he had heard in any speaker'). Its a fact that CAD has helped out horns immensely!! TAD **has** to have known about that problem but over the years/decades they did nothing about it.

Unsound, what do you mean by 'waveform integrity'? Is this in the circuit of the speaker or is this the output, etc and how does this relate to phase angle, if it does?

I'm pretty sure that Duke is trying to get the best fidelity out of his speakers and I have not seen that phrase used before, so I want to be clear about what you mean by it. TIA

Well I can't be the judge of that- I have no idea what the waveform looks like once its in the air. But compared to a pile of speakers that are supposed to be pretty good at that sort of thing I don't find good horns to be lacking.

I find bad horns to suck and not under consideration for HiFi. I don't blame you for not liking horns. I think I mentioned before that they had to win me over, and to this day when I see horns I'm not familiar with I tend to be skeptical.

IOW, Like everything else in high end, its all in the execution.

Macrojack, you had me up until the DSP part. I've spent a lot of time with digital (using both DSP and plain vanilla) and analog recording systems. We use both in our recording studio. While digital falls flat on its face using the master files against the same thing on analog, DSP can only hope to make it worse, in the process of whatever its doing: EQ, compression, decompression, whatever.

I'm happy to keep an open mind but the DSP thing is really going to have to work at it to win me over.

Based on what Duke has imparted I would really like to try a conical horn, combined with a proper field-coil driver and crossover. It really **does** seem as if the potential of horns is still untapped.

Herman, your machine is doing 64bit calculations but the DSP signal is only 24 bit, if that. DSPs can be harnessed to a lot of tasks; one thing you learn really quick in dealing with them is they suck the life right out of the music.

The master tapes in the studio are as close as you will ever get to the real thing. If they can't do it there, they are not going to do it in the home either.

We used a DSP-style crossover at a recent show (T.H.E Show). We were recording live and comparing the live vs the recording. It did sound pretty good, with 24-bit master files, but I am left wondering, how much better would it have been if we had an analog recorder for the recording task, and a passive crossover that allowed a single amp to do the whole speaker rather than a DSP with two very dis-similar amps to do top and bottom?

My experience with electronic crossovers of any type is that they act just like detail filters. In some ways I think the DSP units really do take the analog electronic crossovers to task, but I have yet to see any kind of bi-amped system beat one that is full range.

IOW I am not dismissing DSPs just out of my studio experience, it just seemed like the easiest quick explanation.

Like many other things in audio, its all in the execution. But there is also an 'ignorance is bliss' thing that seems to play out. I've not heard Herman's system so it may well be me who is ignorant.

One objection I've always had to generic electronic crossovers is that they don't allow any tweaking of the curves to accommodate the sort of nuances that Duke has been mentioning. Almost any serious passive crossover you see has tweaks in it to manage idiosyncrasies of the particular drivers in use.

So Herman, why do you use the DSP? Are you bi-amplifying?

You leave out one critical part of your thesis, planers radiate sound from both sides , this ability helps to create the required space and time of a recording far more accurately than any monopole transducer.

Horns will always sound like hi-fi, never real , they project sound in a manner where all instruments and voices have the same projection and size. The best you can say is that they sound just like an amplified concert 40 ft away. Unfortunately we listen to recordings of live music and not live music itself , as such Horns do not convey this as accurately IMO.

If you cannot get the speaker far enough from the rear wall, the first comment is moot. Duke has built systems that are rear-firing and they seem to work fine...

The second comment is simply absurd! Horns image and present soundstage images as well or as badly as any other speaker technology.

With regards to amplifier power and amplifier types used on horns:

My speakers are horn hybrids, using a high efficiency pair of 15" woofers, which limit the efficiency to 98 db.

If I push the system hard, I am really challenged to clip a 30-watt amplifier. But I prefer 150 watts if I can get it, even though I will never use the power. In my case, because I am using OTLs, the amps have no issues at low power: the less power, the less distortion, quite similar to SETs.

The system plays very clean and is devoid of loudness artifacts. The only way you can tell how loud it is playing is if you try to talk to someone sitting beside you or if you have a sound pressure meter.

Amplifiers like the kind I am using (zero feedback) tend to make more distortion following a curve where the distortion becomes more pronounced as you approach clipping. Distortion is where loudness cues come from- the result being that while I can get satisfying volume from the 30-watt amp (and right now I am playing a pair of type 45-based P-P amps that only make 5 watts), the simple fact of the matter is that there are less loudness cues when I use the bigger amps.

(FWIW, the Trio is only horn speaker I know of where the designer intended it to be used with transistors. This is reflected in the crossover design, or lack of it, which consists of capacitors to roll off low frequencies for each driver. This results in an impedance curve that is nearly 19 ohms in the bass horn, but only about 4 ohms at the tweeter frequencies, even though the individual drivers are all nearly the same impedance. No low power tube amp is going to drive this right as the speaker is what I call a Voltage Paradigm technology, whereas most low power tube amps and other horns are Power Paradigm technology. see http://www.atma-sphere.com/papers/paradigm_paper2.html for more info.)

In a case where the speaker is 10 db more efficient than mine, the need for power does get eroded; 15 watts is a lot of power on such a speaker but IMO you do this to reduce loudness artifact from the amp. Horns tend to be very reactive loads and so are often shouty and shrill if the amp used has a low output impedance, particularly if that low output impedance is due to loop feedback in the amp (the back EMF tends to get into the feedback loop, causing the feedback signal to contain false information). This is one reason why horn users who get good results rarely use transistors, and a major reason why many people think that horns are for PA, not hifi. In a nutshell, its an equipment mismatch.

Unsound, Herman and I agree on something :)

Herman, The reason the Trios and probably Duos have such an impedance curve has to do with the way the crossover works. Let's use the Duo as an example, but this is only an example, my experience with the Duo (even though a lot more of our customers use them than Trios) is limited.

So we have the main horn and let's assume that it is 8 ohms. Then there is the tweeter horn and it is 8 ohms too. Now if there is a cap to keep lows out of the tweeter, what happens is that at low frequencies the amp sees the 8 ohms of the main horn. Since there is no choke to keep highs out of the horn, the main horn will be in parallel with the tweeter at tweeter frequencies. So the impedance drops to four ohms because the amp has to drive the main horn and the tweeter too. The Trio uses an expanded 3-way version of this scenario.

If there is a crossover offered that has more than caps in it I think it will be a big step forward for that speaker!

Herman, just to be clear here: I don't know the actual impedance of the drivers in the Duo, but in a case where the minimum impedance is 8 ohms, I would expect most amplifiers will have the voltage response to handle that. Its when you go from 19 or so down to 4 ohms that many smaller tube amps are challenged. So- I think we agree on this one :)

Eldartford, cheesy horns do have resonances, the way many other speaker technologies do. A good horn has less resonance than the your speakers do though. Don't misunderstand me here, I totally empathize with people that have trouble understanding this last point!

As you point out, sound radiates as a spherical wavefront, no doubt why Quad made the ESL63 a semi-spherical device.

A horn often uses a diaphragm that has a similar semi-spherical shape, only it is much larger when it emerges from the mouth of the horn. This may explain why so many people, on hearing good horns, comment that they sound like the best planars, only with greater dynamic impact.

IOW my experience of horns is that planars are really the only thing that compete with horns for naturalness of timbre, coherence and detail- cones don't seem to keep up.
So I rate horns first if they work right (last if they don't), planars second in the firmament and cone systems last.

Weseixas, it sure looks to me like Duke addressed the issues that you had with my last post. I just have one thing to add, and this is true of digital vs analog, tubes vs transistors, horns vs planars, pretty much you name it in audio:

The better the technologies get the more they sound the same and that is simply because the better any technology gets, the more it sounds like real music.

I believe that most horn detractors (which often includes me) are leery of horns due to past bad experiences. However we have to accept that this world offers a large range of experiences, some of which are not available to everyone even if they have been in the field of endeavor for decades. I am now suggesting that while you have heard many horns sound terrible, you have not heard the best horns nor have you heard them with the best supporting equipment that can show them off. I'm going to take it a step further though, to illustrate my comment at the beginning of this paragraph. The same is true for you of planar speakers as well. I am sure that these two statements apply to everyone on this thread.

All we can be really sure of is that the better technologies get, the more they will sound like music and therefore more like each other, not less.

Mapman, for example; singers with 10' mouths.

I'm not getting that at all. I find the imaging to be very pinpoint, as good as any cone or line array.

Weseixas, if you want to quote someone, you use the brackets around the word 'quote'. Once you are done with the quote, you use the brackets again, this time around the word '/quote'

I can even quote myself:

like this

Sometimes its difficult to see what you are saying, since its hard to know if you are quoting someone or not.

Gawdbless, I philosophically disagree with you, as I don't believe it's reasonable to expect speakers to "enhance music". If a speaker accurately portrays a recording for better or worse, I think it very fair to blame the recording in some instances. If one needs to tamper with a recording to make it palatable, I think there are better devices with much more specific controls available to those that feel the need for such things, than speakers. One good example of a recording that that can sound a bit hard is Miles Davis' "Sketches of Spain".

I'm with Unsound on this one, except for the comment about Sketches of Spain. I have original vinyl on that and its got no harshness at all, so long as the system I am playing it on can track it. It *does* have a lot of energy...

The Better any system get's to the truth, is the less you will experience these so called un-listenable tracks.

IME bad recordings bring out the worst in the system as well. So if the system allows you to listen to bad recordings without editorializing, so much the better. I often use bad recordings to give me an idea of hidden artifact in a stereo.

What has been interesting about this is that I have found recordings that I **thought** were bad, but in time were found to have so much energy that a bad system couldn't play it right!

As you might imagine this has nothing to do with horns one way or another.

Unsound, if you can I recommend trying to find a way to hear the original pressing on LP of Sketches. I think you will find that the mic technique is not that bad! I've spent a lot of time in the studio and what you find when you do that is that digital systems lack more than just low level resolution. To make them sound right, you have to accommodate the differences and do the mix differently. I've had to go back and forth in digital mixdowns dealing with this issue: what you hear when you mix is not what you hear in the digital result. Older recordings from the analog era can't benefit from this process, so things that might seem to sound a certain way are found to not be so once you hear the original.

Weseixas, you might want to stop while you're ahead on this one.

My comments was not meant to be derogatory to musicians. Worked with and for many in the past, My comments were nothing more than to differentiate the experience between audiophile and musician ...

-Sounds a lot like: 'some of my best friends are _______ '

IME musicians are very interested in good sound. I play in a band, and although I would not consider the other musicians in my band audiophiles, they are concerned enough about the sound that they bother to do all their recording analog (w/24-bit digital backup). The difference is nothing more than money- if they had it, they would spend it on stereos too.

If you are thinking of Klipsch, it seems the older they are the better they are. But if the price is right, what the heck. They work OK next to the wall. I think I would go for Altecs first as well.

A few years ago I found a set of Electro Voice Regency 3s for cheap. They needed new wiring and new crossover controls. They are designed to use the corner of the room, that is, you put them all the way into the corner. If you don't do that they will not play bass as the cabinet and corner form a sort of horn. Anyway, I just set them up by the Snell Bs I had at the time. They were immediately more spacious, more dynamic. I expected the greater impact, but the increased resolution and depth of the soundstage floored me. Instruments hung in the air, expressed with great timbrel accuracy, delicacy where it was needed and power where needed as well. In short, played every aspect you wanted out of a speaker better than the Snells, except for the bass.

Since I was never able to put them in the corners, I never found out if that was still true when they were properly set up, but this experience let me know in spades that horns had something to offer.

I remember once Harry Pearson (back in the 80s) commenting that horns weren't 'high end'. That sure has changed- no way he would say that now.

Ralph, way wrong. There were great horns when Harry hated them.....Tannoys have been great for a long time, various JBLs were, horns were all over in recording studios sounding great.

Kiddman, I think you may have misunderstood my post. I happen to agree there were great horns when hp made that statement and one of my early customers was making Hartsfield reproductions which sounded great at the audio shows in the late 80s and early 90s. I was running Altecs with dual 15" woofers back in the 1970s... Anyway my point was with my last post that hp would certainly not say that today.

As far as contributions- hp created most of the vernacular we audiophiles use to describe the sound of our stereos- 'soundstage', the use of color descriptions to describe tonality ('bright', 'dark', etc.). He was one of the very first to describe the sound of equipment based on listening. Stereo Review and other magazines around at the time simply did not do that. Nowadays we take that sort of thing for granted. So his contributions should not be ignored even if you don't agree with his reviews. He had a serious influence on high end audio.