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

[Newman]

Well, I thought I had read every paper in Geddes' website but I must have missed his thoughts on MEH's. Do you recall where you saw him say that? I have only heard one MEH, and it sounded wrong to me. I'm not sure if it was that particular design, some other flaw, or an inherent nature of MEH's.

"The whole concept of a Unity horn and a waveguide are at odds. Waveguides seek to propagate singular wavefronts and as such require very specific wavefronts at their throat. The Unity concept does not feed the horn in a manner that is consistent with the waveguide approach." Link

"A waveguide with woofers in the walls will seriously degrade the performance of the HF response. I've designed a lot of the "woofers/mids" in the horn walls types of systems and they never can compete with a stand alone waveguide above the crossover. At crossover they both have problems - neither approach is ideal. I'll take the near ideal HF performance of the solo waveguide any day." Link

Cheers

 

[tmuikku]

You guys need to understand there is different applications for speakers. When your listening height is about static, as in home stereo, then I think it's better not use MEH. You'd anyway get coherent sound because you designed the system so that you are at the "main lobe" so you could optimize things MEH cannot.

But, if you have a large audience, MEH is super. Speakers high up in air and people close to the speakers are at high angle and get reduced SPL and relatively coherent sound. People further back in the audience are more on axis and get similar SPL and coherent sound. It's also relatively easy to deploy, looks cool, gets loud.

When you understand your context, you can find/make optimal system for it. If you don't understand that all loudspeakers are set of compromises optimized for some particular context (and application) you get into silly pointless debates.

 

[gnarly]

Hi tmuikku, you make a good point about the different needs for larger audiences, and a person seated for home listening.
Good sound at every seat, vs optimized for one.

I'd like to talk about my experiences with MEHs in home, compared to a conventional waveguide on top of cones. As you know, I've built a number of different MEH versions.
Before I became enamored with MEH's I built a number of horn/waveguides over cone drivers, and still have those to compare my MEH versions to.
None of my horns/waveguides (I'll just call then horns from here on*) are quite as refined as an OS, but they are all well regarded commercial products.

In assessing and tuning any horn, MEHs included, I first take a set of polars on the spinorama to get an idea what axis makes for a good median reference tuning axis.
Typically, it is not 0 deg H, 0 deg V. Typically it is around 10-15 deg H, and little off 0V depending of how axisymmetric the horn is or isn't.
(I've heard OS guides share this property?)

With a good-compromise reference axis determined, I correct response to be near perfect flat.
Then, i remeasure the corrected response on the spinorama, to see how much the polars deviate from the flat reference .
I view the lack of polar deviations, other than the intentional pattern control design, to be the mark of a both a good horn and a good tuning.

Ok, point of all that.... my commercial horns vs my MEHs...
the commercial horns without any ports like the MEHs, and with better mouth terminations than the MEHs,
inevitably show less polar variation from the flat tuned reference axis.
That said, it's a matter of degree, the MEHs aren't bad by any means.

The tradeoff, in my measurements and listening tests has been...... is the amount of extra cleanliness in the HF/VHF with the commercial horns,
great enough to let go of the extra mid and low cleanliness that I hear in my MEHs.
The MEHs definitely integrate the mids and lows with the HF/VHF better, than my conventional cd&horn boxes, be they DIY or store bought.

Does HF/VHF dominate SQ?
Or is it better to let go of a little HF/VHF perfection, to gain a more optimal full-range integration?
To my ears, the latter... hey, but to each his own !

 

[tmuikku]

Yeah fair enough, I have no MEH available so can't argue. Implementation and suitability for applicatio is key what ever system it is. Of course one can always get bigger than necessary as well, extra headroom and stuff are good bonus, but not necessary.

Even simple full range driver direct radiating system can be pretty nice, very simple and cheap. It's just not enough SPL for some applications like party. Likely measures worse than MEH with taps, top end changes rapidly when listener moves.

 

[Newman]

Quite right. In the end it is all a matter of personal standards vs adaptation effects vs sighted listening effects. Unfortunately it can lead people down, shall I say, rabbit-hole mazes of wrongness, in the belief they are on the right track. The only two ways to get out of the mazes, as far was I have gleaned, are (A) personal controlled listening DBTs of A vs B trials, or (B) building one's objective awareness through the controlled DBT reports of researchers (not just of product vs product preference, but also of human audibility and preference of measured performance levels and imperfections), and applying this knowledge to our choice of audio gear, especially when the right measurements are available. (A) is exceedingly rare but commendable, and (B) is 'the ASR way' and allows much broader application, especially when people like Dr Toole put a life's work into plain-English presentations of (B) for us.

cheers

 

[gnarly]

I'd say there is a (C) way too

(C) continually listen to a bunch of different speakers.

Enjoy them all for what they do well. No problem having favorites...just don't marry any of them.

 

[mhardy6647]

Try page 42

42 is always the answer... to Life, the Universe, and Everything.

 

[dped90]

Do be cautious about driver separation distances with Marcel horns, especially as they get larger.

How much would it help to keep the mouth of the horn closest to the top of my midwoofer?

 

[dped90]

I'm fairly certain he's addressed it in the "Geddes on Waveguides" thread at diyAudio, but I don't have any links saved.

https://www.diyaudio.com/community/threads/is-this-a-horn-or-a-waveguide.178267/post-2597017

If I recall correctly (no absolute guarantees), he doesn't consider the inherent lobing of vertically-aligned drivers in a well-designed two way to be a significant problem in practice, but does consider the extra diffraction/HOMs from the midrange and woofer ports in a MEH to be at least a potential problem.

That's reassuring versus what Newman had cautioned but thanks to both of you for alerting me. Marcel know about mine https://josephcrowe.com/blogs/news/altec-416-8b-in-100l-sealed but had recommended this 20.47" model. https://cults3d.com/en/3d-model/various/ath-a520g2 Will follow up with Earl about it. If he feels it's too big to risk using there's the new A460G2 at 18.11", same size as the A460D.

 

[Newman]

That's reassuring versus what Newman had cautioned but thanks to both of you for alerting me.

No problem. I gather we are talking specifically about driver spacing here. Well, I think that as a general piece of advice, for people who haven't committed to one specific horn design yet, there is a cautionary note to make about driver spacing. For example, I was once working up a concept for a midrange horn, and when I mounted the midrange and HF drivers on a baffle with the required separation and played music, I could clearly hear music coming from each driver separately located in space, one above the other. For my listening distance of 3 to 4 metres, those two drivers were too far apart. I abandoned that particular concept there and then.

The Marcel horn profiles are fairly generously proportioned with large mouth treatments, leading to large-ish driver separation, so I suggest to people working up a horn concept, it is worth considering.

cheers

 

[Newman]

I'd say there is a (C) way too

(C) continually listen to a bunch of different speakers.

That is (A)*, if done properly.

If not done properly, your (C) is a shortcut into the maze while on psychotropic drugs. Exactly what I am saying we need to avoid, if we want to discover what sound waves themselves we actually prefer.


*see post #1045 for descriptions of (A) and (B)

 

[bmc0]

I was once working up a concept for a midrange horn [...] I could clearly hear music coming from each driver separately located in space, one above the other.

What crossover frequency? Elevation localization largely depends on spectral cues above ~3-4kHz and a large horn/waveguide like Marcel's designs would generally be crossed well below that. My DIY main speakers have a center-to-center spacing of 32cm and I sit only about 2.1m away. Despite this, I get zero audible hint of there being two vertically separated sources. The crossover is LR6 acoustic at 1050Hz.

 

[Duke]

Elevation localization largely depends on spectral cues above ~3-4kHz and a large horn/waveguide like Marcel's designs would generally be crossed well below that. My DIY main speakers have a center-to-center spacing of 32cm and I sit only about 2.1m away. Despite this, I get zero audible hint of there being two vertically separated sources. The crossover is LR6 acoustic at 1050Hz.

I believe you. What you describe is similar to my experience.

I would not be surprised if you can sit a little bit closer than 2.1m and still hear zero audible hint of two vertically separated sources, given the specifics of your speaker system.

 

[Newman]

What crossover frequency?

I honestly don't remember.

Elevation localization largely depends on spectral cues above ~3-4kHz and a large horn/waveguide like Marcel's designs would generally be crossed well below that. My DIY main speakers have a center-to-center spacing of 32cm and I sit only about 2.1m away. Despite this, I get zero audible hint of there being two vertically separated sources. The crossover is LR6 acoustic at 1050Hz.

Your 32 cm at 1kHz is basically one wavelength, consistent with the generic suggestion for maximum separation. Now try putting a horn on your mid-bass driver and see what happens.

 

[bmc0]

Now try putting a horn on your mid-bass driver and see what happens.

Even with significantly larger C-C spacing, I don't think one would tend to hear two sources with a ~1kHz crossover given (my understanding of) how localization works.

 

[Holmz]

…
The Marcel horn profiles are fairly generously proportioned with large mouth treatments, …
…

I am seeing “grill” for both mouth and speaker treatments,

 

[MattHooper]

For example, I was once working up a concept for a midrange horn, and when I mounted the midrange and HF drivers on a baffle with the required separation and played music, I could clearly hear music coming from each driver separately located in space, one above the other. For my listening distance of 3 to 4 metres, those two drivers were too far apart. I abandoned that particular concept there and then.

Nice to hear of your experiments making loudspeakers!

Out of curiosity, was that impression derived during personal controlled listening DBTs of A vs B trials?

 

[MattHooper]

(somehow I ended up at the front of this thread and saw this old comment)

Same as yours. I have been around a group of audiophiles that spent untoward amounts of time and money designing esoteric horn systems. Everything they came up with sounded "weird" to me. But they would listen enthralled and try to figure out what the next tweak would be.

This makes me wonder about what people are listening for in our informal listening sessions.

There’s a huge amount of focus, generally speaking in this forum on neutral frequency response as a goal (understandable) and so I wonder if folks like oranegello suffer from some befuddlement somewhat based on this.

In other words, perhaps some people hear some uneven colouration in the frequency response and that’s enough to write the speaker off. It’s just not going to sound right to them.

But another person listening is perhaps zeroing in on other factors that they find compelling. Maybe it’s the dynamics. Maybe it’s something about the frequency response that is making the voices and instruments sound more vivid, alive, dense, palpable or whatever.

I’ve had numerous experiences where an audiophile friend and I listening to the same speakers pretty much perfectly converge in exchanging notes on what we are hearing in terms of the overall character of the speaker.
We agree we seem to be hearing the same characteristics. But my friend will say he likes those characteristics whereas they may irritate me or just leave me cold and I wanna go listen to something else instead.

Reminds me of listening to the modern B&W speakers. I can think of a group of five of us who have listened at length, all of us note the frequency response valleys and mountains, especially the boost in the upper mid range and upper treble range, and the added vividness. But one guy absolutely without reservations loves the sound, I sort of enjoy it as an interesting place to visit, but could never live with those speakers, and some others just find it too much, enough to be irritating. But I certainly get it if you’re looking for a real sense of sound “ not coming from a box” and you love vividness, and hyper detail, those speakers could be your jam.

Similarly, I can see what some people get out of some horn speakers. Like I’ve mentioned before when my friend had the Klipsch La Scalas I wanted to visit and listen to those more often than just about any other speaker he had in his room. They didn’t do much bass, and didn’t as I recall have very airy highs, but man the sense of density in the sound was so compelling - there was a life like factor that many other loudspeakers didn’t match in his room.

 

[gnarly]

Nice to hear of your experiments making loudspeakers!

Out of curiosity, was that impression derived during personal controlled listening DBTs of A vs B trials?

Bingo

 

[Tim Link]

Out of curiosity, was that impression derived during personal controlled listening DBTs of A vs B trials?

Bingo

Maybe it's not something we'd perceive under blind testing. But we don't listen that way, and what we see strongly effects how we perceive the sound even if we know what we see and what the soundwaves are actually doing are different. I'm becoming a wholistic perceptualist these days. Yes, measurements matter. Yes, it's great to know if it's a visual stimuli acting on the auditory perception or if it's strictly soundwaves that are creating the perceptual difference. But perceptions are what we are after. Right? If seeing the distance between the drivers makes it difficult to listen, then you have to do something about it.
I was just thinking about high end gear that I failed double blind tests on so I got rid of it. Wish I hadn't done that. It sounded fine and liked it better in every other respect, and I was enjoying the music more until I got all caught up in the fact that I couldn't actually hear any difference with my unassisted ears alone.
The McGurk effect is just one interesting example