Great Waveguide List

The scale is chosen so that a contour line shows up that corresponds to the beamwidth (-6 dB). If I were to force the PSE-144 sonogram to follow a particular standard, you would not so easily and intuitively see how its dispersion compares to the 80 degree nominal target.



I think it could be worked out with a better use of color (achieving both the dominate beamwidth while still providing considerable detail). I'm not sure if Arta allows this function though, so it would likely have to be corrected in something like photoshop. :eek:


I do think it's a good job of measuring (under domestic constraints) - with very good lower freq. measurements for the horn. :up:



Anyway.. the PSE-144 looks like a good candidate for this sort of design solution :) :

HAIGNER Horn Loudspeakers | Alphahorn
 
Dear Santa Claus,
I want to see both raw and normalized graphics. If on-axis spl is shown, directivity sonogram can be just the normalized version. I'd like to see both 1/3oct and 1/12 or 1/24 smoothing too because I want to see the forest and trees. Impulse gating must be told too, this is very important specially when measurements are done how most of us are usually doing it, behind closed doors. Sometimes I also want to see unsmoothed response measurements of various angles horizontally and vertically, with various gating - sort of hardcore for the late hours...

I will be a good boy!
Juha
 
Scale seems fine to me, it had -6 and -12 dB contour lines.

I also prefer them w/response normalized so as not to muddy the directivity waters.

A single color for everything above -6 dB is pretty coarse. And as to normalized it makes the results look completely wrong with no way to tell what is really going on unless you post the normalization curve as well and then it is very difficult to adjust what you see when you need to look at two different graphs to tell anything valid. You cannot look at a normalized plot and what the frequency response is at any location - you can tell that at every location if they are not normalized. I'm sorry, but normalizing is completely wrong - it is NOT the way the speaker radiates sound.

And as to the DI being different than an Abbey/Summa, there really isn't any difference (if we consider averages). The PSE-144 is nominally 80 degrees and the Abbey/Summa is 90. The control stability in the Abbey/Summa is much better however.

Paul - did you get a license from Danley to produce these? They clearly violate his patent. He probably does not have an AUS patent, but you could not export them to the US without a license.
 
Last edited:
A single color for everything above -6 dB is pretty coarse. And as to normalized it makes the results look completely wrong with no way to tell what is really going on unless you post the normalization curve as well and then it is very difficult to adjust what you see when you need to look at two different graphs to tell anything valid. You cannot look at a normalized plot and what the frequency response is at any location - you can tell that at every location if they are not normalized. I'm sorry, but normalizing is completely wrong - it is NOT the way the speaker radiates sound.

Agreed on one color for a 6 dB range.

When looking at a sonogram I don't want to see the frequency response is at any location; I prefer polars for that.

If normalized you can see directivity characteristics on a soinogram at a glance; otherwise how the heck can you separate the directivity from response variations?
 
When looking at a sonogram I don't want to see the frequency response is at any location
That is exactly what I do want to see.

If normalized you can see directivity characteristics on a soinogram at a glance; otherwise how the heck can you separate the directivity from response variations?

I do not want to separate the response from the directivity. They are intimately coupled in the real world and that's the way I want to see it.
 
Last edited:
Perhaps that useful to you as a skilled designer, but as an end user I'm interested in the directivity characteristics on their own.

Ah, maybe I see what you're getting at - if the response varies as a direct result of spatial power distribution aka directivity, then you'd want to see that.

My original thought was that I wanted the response normalized to take out the effects of the driver.
 
This list is on a discussion forum! I thought that discussion about directivity measurements of waveguides was relevant in this topic, sorry.

Post #1 By Gainphile quoted (notice last row):

Please post your finding and keep the list clean
cool.gif


Link
Driver used
Pics
Price
Where to buy
Polar plots
Experience, subjective observations etc
 
Hate to be a dick.
The high frequency peaks and dips in that measurement are similar to what I ran into when I built my own horns and waveguides.
IMHO, the first five or six inches are really critical, and these waveguides do it better than a plain ol' conical horn:

dsc_4913.jpg

Img_0834.jpg


If you take a look at the frequency response measurements in this thread, you'll see that the unequalized response of most of the elliptical waveguides are superior to conical.

SORRY

I know I'll probably get slammed for this, because that Synergy horn is beautiful, but we can get better results with $14 waveguides.

If you want to get directivity down to 250hz, just extend the JBL waveguide.

IMAGE_3CFE93FF-F235-4D89-B7E5-25CE7BC7E384.jpg

Here's the frequency response of the Danley SH50. See the deep dips at 4khz and 8khz? That's because it's conical.

post-119817-0-08168400-1398985084.jpg

Here's the unequalized response of Mike's Synergy horns. See how it exhibits the same peaks and dips?

Econowave.jpg

Here's the $14 JBL waveguide clone from Pyle. See how flat it is? That's because it's not conical. Conical horns create resonances. Not a big problem at 1khz or 500hz, but at 4khz or 8khz conical horns should be avoided.

LeCleach did some measurements of conical horns in France and found the same thing.
 
Last edited:
John

I think that there is much truth in what you say, but I also think that it is over simplified. The conical flare alone is not necessarily at fault. It is the treatment at the mouth and at the throat. In the synergy horn neither of these locations are done properly for minimum diffraction and the resulting reflections. This aspect is at least part of the problems that you describe.

The synergy concept is great for very large system required by big venues with very high power outputs, but I don't think it is a good choice for small rooms or home theaters (or even commercial theaters for that matter). It was designed for stadiums and I am sure that in that application it works quite well.
 
John

I think that there is much truth in what you say, but I also think that it is over simplified. The conical flare alone is not necessarily at fault. It is the treatment at the mouth and at the throat. In the synergy horn neither of these locations are done properly for minimum diffraction and the resulting reflections. This aspect is at least part of the problems that you describe.

The synergy concept is great for very large system required by big venues with very high power outputs, but I don't think it is a good choice for small rooms or home theaters (or even commercial theaters for that matter). It was designed for stadiums and I am sure that in that application it works quite well.

Eight years ago I wanted to make some waveguides.
I'd read the things that you'd published, so I settled on an oblate spheroidal curve and a coverage angle of ninety degrees.

I went through all the time consuming steps required to make one. I wrote a spreadsheet, I built a mold layer by time consuming layer. When I was finished, this is what it looked like:

qsc-vs-osWaveguide2.jpg

qsc-vs-os-waveguide1.jpg


My mold in the front, QSC PL-000446GP behind it.

As you can see, they have a lot in common. And this was completely accidental - I made my mold based on what you wrote, a solid three years before the QSC was available to us.

Picture007.jpg

The JBL PT Waveguide? It's almost a dean ringer for the QSC. And for mine. Which was based on what you wrote.

Basically all of these waveguide designers are 'zeroing' in, and all of our waveguides are starting to look a lot alike. Here's what they have in common:

1) They're asymmetrical. Asymmetry reduces peaks and dips in the frequency response. Yes, you can move the peaks and dips off axis, like you did in the Summa, but asymmetry works too.
Picture008.jpg

2) They're oblate spheroidal or close to it. A significant fraction of the waveguides in this thread meet that criteria. The JBL PT waveguide, your waveguides, the QSC waveguides, the SEOS waveguides, the waveguide Wayne Parham uses. In the pic above, the Pyle horn at the top left doesn't meet the criteria, the JBL and the QSC in the pic do.
3) Wide coverage angles work better in a home than narrow.



attachment.php

One thing that's hard to appreciate is that the difference between conical and oblate spheroidal is quite small. It's a subtle change at the throat and the mouth. You already know this of course :)

An externally hosted image should be here but it was not working when we last tested it.

But I want to bring this up because a lot of people would look at this and think 'what's the difference?'

The OS curve is soooo close to conical, but the difference in sound is unmistakable. Here's an anecdotal example of this:

I've owned Summas for six years now. Last year I started to study LeCleach, and I realized that LeCleach horns DO improve some aspects of horn performance. I built a LeCleach horn (badly), and I also bought some exponential horns. I had the Summas in my bedroom, the other horns in my living room. And I found that I just had no interest in listening to the other horns for more than about 10-20 minutes at a time. This was remarkable, because the horns *definitely* did a lot of things right. They had an 'x-ray' quality that was compelling, the decay of musical notes sounded different on the horns. (Superior group delay is probably the explanation.) But I just couldn't listen to them for long. They were fatiguing.

I had a pile of the Pyle clones of the JBL PT waveguide laying around. I bought them just to see what all the buzz was about. I had no intention of using them, literally bought them out of curiosity.

I hooked up the same compression driver to the Pyle PH-612, and suddenly found myself listening to music again. No, it's not an xray like the exponential horn is, no you're not going to extract every last bit of detail. The PH-612 basically sounds a lot like a regular ol' dome tweeter. Just louder, with lower distortion, and better directivity.
 
John

I think that there is much truth in what you say, but I also think that it is over simplified. The conical flare alone is not necessarily at fault. It is the treatment at the mouth and at the throat. In the synergy horn neither of these locations are done properly for minimum diffraction and the resulting reflections. This aspect is at least part of the problems that you describe.

The synergy concept is great for very large system required by big venues with very high power outputs, but I don't think it is a good choice for small rooms or home theaters (or even commercial theaters for that matter). It was designed for stadiums and I am sure that in that application it works quite well.

You have valid concerns with the Synergy horn. When building a Synergy horn, there are three big differences between your waveguides and Tom's horns:

1) The Synergy horns are conical horns with treatment at the throat and the mouth; the Gedlee waveguides are oblate spheroidal
2) The Synergy horns have eight holes in them, for the midrange and woofer taps. These undoubtedly impact the frequency response negatively and they probably generate higher order modes
3) The phase response and the step response of the Synergy horns is superior to the Gedlee speakers, because of the crossover and the tight coupling of the drivers

But here's the thing:
Nearly all of the things that are good about your design can be applied to the Danley design.

monster-massive14.jpg

monster-massive31.jpg

Here's a graph comparing the polar response of QSC's waveguide, which is as good as they come. That QSC is a fantastic waveguide. The second graph is the same waveguide, but with cardioid midranges bolted to it. The second graph is definitely lumpier than the first.

IMAGE_3CFE93FF-F235-4D89-B7E5-25CE7BC7E384.jpg

With some careful treatment of the midrange taps, and a very good waveguide like the QSC, I think it's possible to 'flatten out' the top three octaves of Danley's Synergy Horn.

Bill Waslo and Paul Spencer have done some great work with this too. It's really exciting to see people take your ideas and Danley's ideas and push the envelope.
 
3) The phase response and the step response of the Synergy horns is superior to the Gedlee speakers, because of the crossover and the tight coupling of the drivers

I wouldn't agree with this - it depends. On-axis the close coupling is not an issue in my designs. Nor is it an issue in the horizontal in general. Only in the vertical, at crossover, will the closeness of the synergy drivers be an asset.

It is all about "scaling" up. As the system gets bigger and bigger the distances between the drivers get larger and larger and you need to do something to get them closer together - ala synergy. In a smaller system this is not really a big issue, but then power handling is. The sweet-spot IMO is a two way with high power drivers. One can get plenty of power - as your Summas have - and very good polar control, without any lobing problems.

I continue to dislike rectangular mouths and don't understand why people do that. If you look at the directivity in the plane containing the corners it is terrible. A round or elliptical waveguide doesn't do that.
 
I wouldn't agree with this - it depends. On-axis the close coupling is not an issue in my designs. Nor is it an issue in the horizontal in general. Only in the vertical, at crossover, will the closeness of the synergy drivers be an asset.

It is all about "scaling" up. As the system gets bigger and bigger the distances between the drivers get larger and larger and you need to do something to get them closer together - ala synergy. In a smaller system this is not really a big issue, but then power handling is. The sweet-spot IMO is a two way with high power drivers. One can get plenty of power - as your Summas have - and very good polar control, without any lobing problems.

I continue to dislike rectangular mouths and don't understand why people do that. If you look at the directivity in the plane containing the corners it is terrible. A round or elliptical waveguide doesn't do that.

The phase response from 500hz to 2khz isn't flat; it can't be flat.
Whether that is an issue depends on whether you think phase matters or not.
For instance, I have some JBL "Control Now" speakers here, and their frequency response is very well behaved. Flatter than my Synergy horn experiments. But the Synergies have better articulation and they image better. Obviously, phase is only one of many variables. Is the flat phase contributing to their imaging? The jury is still out.