Horn Speakers - Is it me or.......

[March Audio]

OK, it happened again. I recently listened to some large horn speakers. I have periodically done this over many years at various hifi shows. I came away again with the same thoughts. I just don't get them.

Firstly they were clearly highly directional. Unless I was right bang in the firing line on axis they had no high frequencies.

Secondly, mid range seemed over emphasised with a "cuppy" effect. Exactly like the sound you get if you cup your hands around your mouth.

Lastly they were no more dynamic than any other large speaker.

All the same characteristics I have heard previously.

Is it me? Am I biased? Heard the wrong horns? Some rave about horns but it's lost on me.

What are others experiences?

 

[solderdude]

I heard the speaker shown below on 2 occasions. (in my younger years)
(Will most likely have been the mkI)

The first time I heard it the HiFi shop I worked from a small 10W amplifier.
It sounded really great.

A few months later I heard it at an audio-fair in a large room from a high power amplifier in an unsuited room and driven very loud and hated the sound.

Same speaker, different circumstances... wonderful and terrible.

 

[March Audio]

I heard the speaker shown below on 2 occasions. (in my younger years)

The first time I heard it the HiFi shop I worked from a small 10W amplifier.
It sounded really great.

A few months later I heard it at an audio-fair in a large room from a high power amplifier in an unsuited room and driven very loud and hated the sound.

Same speaker, different circumstances... wonderful and terrible.

That's a reasonable point to make, yes rooms don't always help, but this is a consistent impression I have had over different speakers, rooms etc.

 

[oivavoi]

Subjective, anecdotal experiences: Some of the speakers which have impressed me the most have been horns, and many of the speakers I've liked the least have also been horns. I do think that large horns can present a dynamic "attack" which other speakers can't reproduce. Maybe it's a sort of distortion or higher order modes, I don't know, or it might be about how horn speakers couple sound waves to the air. But there's something there, in any case. But many horns - and even a lot of waveguide speakers - nevertheless sound colored and resonant to me.

All the horn setups I've heard which I've really liked have therefore been DSP based, with generous use of EQ. Unfortunately that's the case for only a minority of horn setups worldwide, as most horn manufacturers and horn audiophiles have a thing for tubes etc. So chances are that when you hear a horn it will sound bad, given that it will probably not be eq'ed.

But they are all highly directional, so sweet spot is limited, unless one uses corner horns which spray the whole room with sound evenly. I also think this directional quality can sound very impressive, but ultimately feels less natural in long-term listening. Having said that, if I had a really big house where I could install two or three hifi systems, I have no doubt that one of them would have been a system with some really large horns.

 

[dc655321]

The only "serious" horns I've heard were an Avantgarde Trio XD system.
This was in a large hotel suite at Axpona 2018 and a remaster of Benny Goodman's "Sing, Sing, Sing" was playing (loudly!).
IIRC, there was about $150k of gear in that room. Make of that what you will.

The old-time-y vibe of the music with the old-time-y visuals of the large horns created smiles all around the room.
I thought they sounded very good, the percussive elements of that song were rendered with great authority.
Still, I got they impression they were more of a "look at me!" than "listen to me!" statement.
Not for me, but I can appreciate that some may be enamored with them.

For a great treatment of all things "horn", the Geddes on Waveguides thread at diyAudio is a classic.
It's extremely long, and wanders about, but even the 1st page has good insights.
Another interesting development on diyAudio lately, in regard to horns, is the release of software for waveguide modeling.
I can dig that link up if any are interested.

 

[fredoamigo]

I don't know if I understood the earl geddes white paper on directivities but it concludes that the performance of a waveguide is incomparable and superior to other configurations? . I think the concern of waveguides is that there are many bad ones and not many good ones .


http://www.gedlee.com/Papers/directivity.pdf

 

[March Audio]

Well these horns beamed the high frequencies like a laser. I can't see anything good about that waveguide

 

[Ron Texas]

Horns are for horny audiophiles.

 

[UliBru]

From my point of view a horn should not cover more than 2 octaves frequency range (hypothesis). Then it does not beam too much.
Furthermore a horn diameter (circumference) should match too.
With most classical horn systems the horn throats have different lengths and thus the timing between the horns is quite lousy. So it is worth to time-align the sound arrival.
I run a fully active 4-way system with three horns 400Hz-1600Hz, 1600Hz-6400Hz, 6400Hz-24kHz. It sounds amazing with a distortion reduced by about -20dB compared to other speakers. Unfortunately the bass up to 400Hz is played by corner woofers (mimicking a horn by the room corners). I would prefer to run a 5-way horn system from 25 Hz-100Hz and 100Hz-400Hz but good horns for this frequency range get incedible large.
Crossovers and time alignment is digitally carried out.
Anyway I'm happy. There is no real alternative.

BR
Uli

 

[mitchco]

Earl's paper on directivity is an excellent one. The issue with most "horns" are that they are poorly designed, as Earl's paper points out. I.e. uneven directivity over frequency and narrowing at high frequencies to a pencil beam width. Most horns vertical directivity is much narrower than horizontal directivity with typical patterns like 90 degrees wide and 40 or 50 degrees vertical.

However, modern waveguides like what is used in the JBL M2 with it's "constant directivity" waveguide with a pattern of 110 x 100 does not have narrowing at high frequencies or any "horn honk" whatsoever. Have a look at the "spinorama" to see what I mean. If you can find a pair (or other JBL's with the image control waveguide) to listen to, you will hear it first hand. Earl's waveguides are also constant directivity using an oblate spheroid and if you look at Earl's whitepaper, you can also see a Polar Map of his waveguide, which is pretty dang good. Note that constant directivity waveguides also require a special type of eq compensation.

There has been a lot of trickle down tech from the pro field into consumer loudspeakers and you will notice that D&D 8c and even the Revel Salon 2 have their tweeters mounted in a "waveguide" for more of a constant directivity control as their polar plots and spinoramas reveal. Both these products have excellent, constant off-axis frequency response at the mid and high frequencies. Almost all pro monitors also have the tweeter loaded in a waveguide...

Unfortunately, the Avant Garde system mention above, along with the pic of the "Smith" horn and bullet tweeter are not constant directivity devices (aside from mismatching directivity between devices). Having heard both in the past, they are not a good representation of what a properly designed constant directivity system should sound like. In fact, the whole idea is that the off-axis response is as good as on the on axis response. My JBL 4722's use a constant directvity waveguide and I get an even mid to high frequency response across my 3 seat, 6 foot couch. No beaming here, but it is an older design compared to the M2. The math is really tricky to design a proper constant directivity horn/waveguide with excellent on and off axis response. But there is some good software at this thread that does it right: https://www.diyaudio.com/forums/multi-way/338806-acoustic-horn-design-easy-ath4.html

@March Audio hope that helps a bit.

 

[March Audio]

From my point of view a horn should not cover more than 2 octaves frequency range (hypothesis). Then it does not beam too much.
Furthermore a horn diameter (circumference) should match too.
With most classical horn systems the horn throats have different lengths and thus the timing between the horns is quite lousy. So it is worth to time-align the sound arrival.
I run a fully active 4-way system with three horns 400Hz-1600Hz, 1600Hz-6400Hz, 6400Hz-24kHz. It sounds amazing with a distortion reduced by about -20dB compared to other speakers. Unfortunately the bass up to 400Hz is played by corner woofers (mimicking a horn by the room corners). I would prefer to run a 5-way horn system from 25 Hz-100Hz and 100Hz-400Hz but good horns for this frequency range get incedible large.
Crossovers and time alignment is digitally carried out.
Anyway I'm happy. There is no real alternative.

View attachment 37370

BR
Uli

So you have to get pretty extreme to get them to function well

 

[March Audio]

Earl's paper on directivity is an excellent one. The issue with most "horns" are that they are poorly designed, as Earl's paper points out. I.e. uneven directivity over frequency and narrowing at high frequencies to a pencil beam width. Most horns vertical directivity is much narrower than horizontal directivity with typical patterns like 90 degrees wide and 40 or 50 degrees vertical.

However, modern waveguides like what is used in the JBL M2 with it's "constant directivity" waveguide with a pattern of 110 x 100 does not have narrowing at high frequencies or any "horn honk" whatsoever. Have a look at the "spinorama" to see what I mean. If you can find a pair (or other JBL's with the image control waveguide) to listen to, you will hear it first hand. Earl's waveguides are also constant directivity using an oblate spheroid and if you look at Earl's whitepaper, you can also see a Polar Map of his waveguide, which is pretty dang good. Note that constant directivity waveguides also require a special type of eq compensation.

There has been a lot of trickle down tech from the pro field into consumer loudspeakers and you will notice that D&D 8c and even the Revel Salon 2 have their tweeters mounted in a "waveguide" for more of a constant directivity control as their polar plots and spinoramas reveal. Both these products have excellent, constant off-axis frequency response at the mid and high frequencies. Almost all pro monitors also have the tweeter loaded in a waveguide...

Unfortunately, the Avant Garde system mention above, along with the pic of the "Smith" horn and bullet tweeter are not constant directivity devices (aside from mismatching directivity between devices). Having heard both in the past, they are not a good representation of what a properly designed constant directivity system should sound like. In fact, the whole idea is that the off-axis response is as good as on the on axis response. My JBL 4722's use a constant directvity waveguide and I get an even mid to high frequency response across my 3 seat, 6 foot couch. No beaming here, but it is an older design compared to the M2. The math is really tricky to design a proper constant directivity horn/waveguide with excellent on and off axis response. But there is some good software at this thread that does it right: https://www.diyaudio.com/forums/multi-way/338806-acoustic-horn-design-easy-ath4.html

@March Audio hope that helps a bit.

Thanks. I have to say I view a waveguide somewhat differently a full blown large horn as pictured above. They are obviously achieving quite different results.

 

[oivavoi]

From my point of view a horn should not cover more than 2 octaves frequency range (hypothesis). Then it does not beam too much.
Furthermore a horn diameter (circumference) should match too.
With most classical horn systems the horn throats have different lengths and thus the timing between the horns is quite lousy. So it is worth to time-align the sound arrival.
I run a fully active 4-way system with three horns 400Hz-1600Hz, 1600Hz-6400Hz, 6400Hz-24kHz. It sounds amazing with a distortion reduced by about -20dB compared to other speakers. Unfortunately the bass up to 400Hz is played by corner woofers (mimicking a horn by the room corners). I would prefer to run a 5-way horn system from 25 Hz-100Hz and 100Hz-400Hz but good horns for this frequency range get incedible large.
Crossovers and time alignment is digitally carried out.
Anyway I'm happy. There is no real alternative.

View attachment 37370

BR
Uli

This system looks... well, pretty, pretty good, I must say!

May I ask why you chose that shape on the horns? I've understood that there's been quite a lot of debate about different types and shapes of horns and which ones perform better, but I find this debate somewhat difficult to comprehend.

 

[Juhazi]

I dont like horn speakers either. Imaging is super, but I get exhausted soon with the feeling of wearing headphones! The dominance of direct sound makes that, but I know many hifi friends who love horns/synergys. "Normal size" 2-3-way speakers with waveguide-tweeter are often the best solution. Bigger waveguides like in JBL M2 are more like horns, but not as beaming as GedLee speakers or "real horns"

Fortunately we have alternatives! One should learn to know these differencies that are based on acoustic principles, and also to understand how different rooms can be. Often hifi shows are in large meeting rooms that have high RT, and then horns have an advantage - sound at home might be an unpleasant surprise.

 

[UliBru]

This system looks... well, pretty, pretty good, I must say!

May I ask why you chose that shape on the horns? I've understood that there's been quite a lot of debate about different types and shapes of horns and which ones perform better, but I find this debate somewhat difficult to comprehend.

Of course you can find a lot of discussions about horn shapes. To be honest I have used these spherical horns as they are not too expensive see e.g. https://eckhorn.com/jabo-kh-72.html or http://horns-diy.pl/en/). And spherical horns behave quite well.

 

[Ron Texas]

@UliBru even hornless, I love my finished attic and it's funky ceiling.

 

[Frank Dernie]

I have horns, normal(ish) box speakers and DSP equalised speakers in here. The horns aren't cuppy and the directivity of the tweeter doesn't seem to be more than I am used to but the mid horn is. The distortion is low but the waterfall is slow decay.
As is usually the case I think good engineering is the answer not speaker type, there are good and bad of any design concept.
When I first heard these horns I thought they could replace everything else but my mains speakers are probably unsaleable and I still like them so I kept them too.
The DSP ones I bought fascinated by the technology but rarely use them since they need connecting to a computer, or mobile, to control volume, which I hate with a passion. They sound great though.

 

[Mtbf]

Crossovers and time alignment is digitally carried out.

Could you elaborate a bit on what hardware you use for the digital crossovers, and how you integrate the subwoofers?

 

[Purité Audio]

I don’t find horns more dynamic, more coloured certainly.
Keith

 

[ahofer]

From my point of view a horn should not cover more than 2 octaves frequency range (hypothesis). Then it does not beam too much.
Furthermore a horn diameter (circumference) should match too.
With most classical horn systems the horn throats have different lengths and thus the timing between the horns is quite lousy. So it is worth to time-align the sound arrival.
I run a fully active 4-way system with three horns 400Hz-1600Hz, 1600Hz-6400Hz, 6400Hz-24kHz. It sounds amazing with a distortion reduced by about -20dB compared to other speakers. Unfortunately the bass up to 400Hz is played by corner woofers (mimicking a horn by the room corners). I would prefer to run a 5-way horn system from 25 Hz-100Hz and 100Hz-400Hz but good horns for this frequency range get incedible large.
Crossovers and time alignment is digitally carried out.
Anyway I'm happy. There is no real alternative.

View attachment 37370

BR
Uli

These types of systems fascinate me, and I’ve never heard one. So many amps and channels to coordinate. Do you find you have to use the same amp for each driver? I would think you could develop some difficult phase and directivity interactions.