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

[mhardy6647]

Who records stereo or multich with "1-mic 1-track 1-channel"??? That's called mono

Indeed. It's very challenging to record mono thus.
Like Dilbert and his database:

https://dilbert.com/strip/1992-09-08

 

[MakeMineVinyl]

I'm having a much harder time thinking of speakers with "generic megaphone shape" tbh

Do any examples come to mind?

Perhaps this?

 

[Frank Dernie]

I have a pair of Genelecs with the mid cone surrounding the tweeter

Yes the 8260 has a horn loaded tweeter. I did consider getting some a few years ago but they stopped making it.

 

[Floyd Toole]

In my 50+ years of loudspeaker research I have heard and evaluated just about everything. In the early to middle years horns were, to my ears, easily recognizable by virtue of their audible "throat" colorations - heard within a very narrow high frequency beam. When I designed my first studio monitor loudspeaker in 1978 (described in Section 18.3.1 in the 3rd edition of my book) I evaluated numerous horns and drivers, ending up selecting a JBL "Smith" horn as being the most neutral sounding. It had almost no "throat". Over the years improvements were made to the directivity control of horns, but this feature persisted. Then a new wrinkle was introduced: short-throat horns. These were a change from the long-throw traditional sound reinforcement designs. For monitor speakers with less output demands and less restrictive directivity control, they seemed to sound better. Certainly they offered less opportunity for air non-linearity at high SPLs. In my experience the real breakthrough came with the JBL Professional M2, with a new horn and a new compression driver. The horn as can be seen is very shallow, with almost no throat in the classic sense. This combined with the new compression driver, which lacks the traditional large dome diaphragm and exhibits extended HF bandwidth was, to my ears, and others, exceptionally transparent. Since then it has held its own when compared to the best of the cone/dome designs. If it were not for the "industrial" appearance, I might well have had them in my recently renovated home theater. The more recent 7 series and 3 series monitor loudspeakers share the newly evolved design features. Consumer versions are appearing.
There are other short-throat designs in the market now, and other designs I have not experienced, so perhaps it is time to retire some of the traditional biases.

 

[stevenswall]

Yes the 8260 has a horn loaded tweeter. I did consider getting some a few years ago but they stopped making it.

I think Waveguide is more accurate than horn, just like I'd say "horn" instead of waveguide if something looks more like this:

https://www.bhphotovideo.com/c/product/853203-REG/pyle_pro_pmp53in_50w_professional_piezo.html

And less like a Revel, Genelec, or JBL waveguide.

Avantguard comes to mind with the flared megaphone horn designs that in some cases look like the megaphone above, or a siren, or an old gramaphone horn.

Seems like most "horns" or rather 'correctly calculated waveguides that aim for an achieve a flat frequency response' that work, are much more shallow and don't look like the ones shown below. Maybe there are companies that measure things and design them as well as genelec, I've just never come across measurements to show that with a "horn" speaker which I'm defining as an instrument or siren horn shape, not a relatively shallow waveguide.

 

[Duke]

I see a horn as a large and deep waveguide. I'll bet I'd like "horns" that are calculated to evenly project all frequencies in a direction without emphasizing specific pitches. I'll bet I'd dislike horns that are slapped on to concentrate the sound and provide 'dynamics' which just means more volume to scratch my ears off.)

Imo waveguides or waveguide-style horns (aka "shallow horns") meet your criteria, as they tend to do the "constant directivity" thing well with no sound-quality downside.

I am curious if there is a way to identify horns that were modeled vs a generic megaphone shape though.

Imo you may want to avoid diffraction horns - those which use an abrupt angular discontinuity to increase the high-frequency coverage pattern. You can usually see the diffraction slots and/or wall kinks by eyeballing the horn. Horns which aggressively use diffraction can sound harsh as the volume level goes up.

Beyond that, at the risk of over-generalizing, the coverage angle of the short wavelengths is controlled way back in or near the throat, and as the wavelengths become longer the coverage angle is progressively controlled closer and closer to the mouth. Many round horns have a curvature which starts out gradual at the throat, and then increases as we get closer to the mouth. Exponential, tractrix, spherical, Le Cleac'h, etc. So the radiation pattern is narrow at high frequencies (because the angle is narrow near the throat), and progressively widens as we go down in frequency because the horn's angle is expanding towards the mouth.

The type of waveguide-style horn I prefer is the Oblate Spheroid (introduced by Earl Geddes), whose curvature mostly happens very close to the throat, and beyond there the walls are almost straight-sided. A large round-over has to be blended onto the mouth to give a low-diffraction termination. The oblate spheroid curve is mathematically the shape which causes the least disturbance to the waveform for a given change in angular coverage, or at least that's my understanding. There are several off-the-shelf horns which imo are also well-behaved, 18Sound and DIY Sound Group's SEOS horns come to mind, look for constant directivity without sharp-edged diffractive features.

The math of the horn JBL uses in the M2 is way above my pay grade, but it LOOKS to me like the "knuckles" near the throat help control the radiation pattern at high frequencies, and then the internal geometry which results in the "valleys" along the diagonals would send any internal reflections off in directions which miss the primary listening area. But this is just a guess. It certainly has no horn signature and is non-fatiguing to listen to.

Maybe there are companies that measure things and design them as well as genelec, I've just never come across measurements to show that with a "horn" speaker which I'm defining as an instrument or siren horn shape, not a relatively shallow waveguide.

Autotech shows polar maps for some of their horn shapes, and they make a fairly wide variety, so you can get some feel for the correlation between what the horn looks like and how it measures. You have to just click on the different individual horns and see whether they have the data:

Horns by Auto-Tech – acoustic horns, waveguides, casings and ready-made loudspeaker systems (horns-diy.pl)

And in case you haven't seen it, here's the data for a design with a large Oblate Spheroid waveguide:

GedLee LLC

 

[andreasmaaan]

Seems like most "horns" or rather 'correctly calculated waveguides that aim for an achieve a flat frequency response' that work, are much more shallow and don't look like the ones shown below.

Shallow vs deep is certainly one parameter with which you can characterise horns/waveguides, but when it comes to the behaviour of the horn, there are other parameters that are more fundamental.

You might be surprised to hear that the "megaphone-style" (exponential, I presume you mean) horns are actually one of the few types that aim to achieve a flat frequency response (above the so-called "cut-off"). The shallower, constant-directivity horns like those found in Revel and Genelec speakers do not produce flat on-axis response without equalisation (e.g. in the crossover).

Maybe there are companies that measure things and design them as well as genelec, I've just never come across measurements to show that with a "horn" speaker which I'm defining as an instrument or siren horn shape, not a relatively shallow waveguide.

It seems to me your nomenclature is a bit confused. What you are calling "horns" seems to include both conical and exponential types. The devices you are calling "waveguides" are primarily conical. The key differentiator therefore seems to be simply the depth of the device, which as mentioned is one of many factors that determines how the horn will perform, but certainly not the most fundamental one.

If you'd like to delve into this topic a little deeper, Bjorn Kolbrek is a great place to start. His two aritcles for AudioXpress are both concise and quite comprehensive. You can find them here.

 

[Floyd Toole]

Dr. Toole, can you relate what you use (briefly...since it is a bit OT) to synthesize stereo to multichannel?

Currently I am using Auro3D as an upmixer as it is the most pleasant of the offerings in my SDP75 processor. Some of the parameters and its overall level of contribution can be adjusted which is useful. However, my expectations are not high, looking only for a slight enhancement of spaciousness and envelopment. It will never replace true multichannel recordings.
Looking back, the best upmixer I ever had was in a Lexicon processor - Logic 7, developed by David Griesinger. It also could be adjusted, as I recall, and at its best it pretty much left the soundstage alone, adding only a gratifying envelopment. I liked it, but it is long gone, sadly.
Turning it off, reverting to stereo, could be truly depressing.

 

[Sal1950]

 

[mhardy6647]

my idea of a waveguide (FWIW).

Just like ol' John Karlson...

http://www.hifilit.com/Karlson/Karlson.htm

https://www.diyaudio.com/forums/full-range/256176-real-inspiration-karlson-coupler.html
http://www.troelsgravesen.dk/download/1955-Karlson-speaker-brochure.pdf

 

[MakeMineVinyl]

Karlson enclosures are something I've only experienced once and to be truthful I don't remember what I thought of the sound. An old 1950s "High Fidelity" review said something like they sounded muted, like the strings had applied their mutes, but that the enclosure would 'take the edge' off of harsh sounding drivers. Here is the link to the archive of High Fidelity magazines. I think it was in the vicinity of 1954 or thereabouts.

It would be nice to be able to time travel and step into this 1957 showroom.

 

[Juhazi]

https://audio.com.pl/testy/stereo/kolumny-glosnikowe/3309-avantgarde-acoustic-uno-fino-edition

Not the best Ava I guess. Step response not shown...

https://www.stereophile.com/content/avantgarde-acoustic-uno-nano-loudspeaker-measurements
https://www.stereophile.com/content/avantgarde-acoustic-duo-loudspeaker-measurements

Extremely low distortion and controlled directivity give the sound. With a little help from dsp-eq and delay result could be much better. Multihorns with multi-channel dsp would be my choice for really big rooms (halls) with solid walls.

 

[thewas]

Looking back, the best upmixer I ever had was in a Lexicon processor - Logic 7, developed by David Griesinger. It also could be adjusted, as I recall, and at its best it pretty much left the soundstage alone, adding only a gratifying envelopment. I liked it, but it is long gone, sadly.
Turning it off, reverting to stereo, could be truly depressing.

Funnily my current daily drive car has also a Harman Logic 7 processor (of course much more simple and not adjustable) and I also prefer it always switched on vs. off which is actually a first time for me as before I never liked any such sound processing.

 

[tuga]

Yeah, I have a pair of Genelecs with the mid cone surrounding the tweeter, and a pair of Kalis, though they don't have much of a waveguide after the midrange cone.

The difference between waveguided speakers and "real" horns is that the directivity of the former becomes narrower as you move up the frequency range from 300-500Hz upwards whilst a "real" horn speaker will have constant narrow directivity across most of the spectrum:

https://www.fidelity-online.de/avantgarde-acoustic-uno-xd-messungen/


In my view it would help if we only called horn to a waveguide which narrows directivity and increases efficiency (a horn is a waveguide but some waveguides are not horns).
In that sense the JBL M2 does not use a horn but a waveguide.
But there is much discussion about this in the JTR topic already...

Quasi-fullrange "real" horn speakers will produce narrow directivity down to the midbass or lower but need to be at least 4-way and tend to be massive and very expensive.

In my 50+ years of loudspeaker research I have heard and evaluated just about everything. In the early to middle years horns were, to my ears, easily recognizable by virtue of their audible "throat" colorations - heard within a very narrow high frequency beam.

In my admitedly very small knowledge and limited experience this "throaty" sound results from crossing the driver too close to the resonant frequency of the horn and that can easily be avoided by raising the cutoff frequency of the lower knee and using a narrower-range bandpass.

http://education.lenardaudio.com/en/07_horns.html

If you add that the horn will lose efficiency some 3 octaves above that, you end up with a 3- or 4-way speaker and a bass bin like the Cessaros below:

Adding enough ways will also eliminate the off-axis flare see above in the 2-way + bass-bin Avangarde Uno (at around 3kHz).
The ilustration below gives an idea on the size of a Tractrix horn for a given resonance frequency:

 

[tuga]

Who records stereo or multich with "1-mic 1-track 1-channel"??? That's called mono

I must have expressed myself poorly.
I meant that for 2- or 5-channel stereo you record using 1-mic to 1-track to 1-channel – no mixing, no pan-potting, only EQ.

If you haven't already then give a good listen to Mario/PlayClassics' "real-stereo" recordings. They are some of the most realistic I have ever listened to:

https://www.audiosciencereview.com/...lassics-album-premiere-for-asr-members.11224/

 

[tuga]

Funnily my current daily drive car has also a Harman Logic 7 processor (of course much more simple and not adjustable) and I also prefer it always switched on vs. off which is actually a first time for me as before I never liked any such sound processing.

A car is a terrible listening environment. It's noisy (engine, aerodynamics, tires, traffic, ventilation), the listeners are off-center and the driver needs to focus on his driving.
It probaly makes sense to use DSP.

 

[thewas]

A car is a terrible listening environment. It's noisy (engine, aerodynamics, tires, traffic, ventilation), the listeners are off-center and the driver needs to focus on his driving.
It probaly makes sense to use DSP.

Yes, but the Logic 7 does more than just correcting the image location for the driver by adding some delays like other systems do and just some "room correction" but increases the envelopment by adding some processes cues to the several surround channels.

 

[thewas]

In my view it would help if we only called horn to a waveguide which narrows directivity and increases efficiency (a horn is a waveguide but some waveguides are not horns).

Actually also most waveguides increase efficiency at least at some frequency regions so the transition is fluent and not clear imho.

 

[Wombat]

Whatever happens in a car has to be a BIG COMPROMISE compared to a listening space in most homes.

 

[andreasmaaan]

Yes, but the Logic 7 does more than just correcting the image location for the driver by adding some delays like other systems do and just some "room correction" but increases the envelopment by adding some processes cues to the several surround channels.

Do you know of any detailed info on Logic 7? E.g. whether it uses a form of LMS-based processing, etc.