doug20 said:
The progressive transition waveguide sounds like a horn because it's a horn.
That's why I compared it to the JBL MP415, which uses a waveguide. "Progressive Transition Waveguide" is market-speak for "almost a waveguide but not quite."
That's why I explicitly asked Xpert and Zilch if they'd listened to a commercial implementation of a speaker with a waveguide. I also listed a few examples of speakers that use them (correctly.)
The function of a waveguide is to control directivity. To do so, the coverage angle must remain constant. That's a fancy way of saying "the walls are flat." There's an exception to this rule, I'll get to that in a second.
Here's a few profiles and pics of waveguides. Note the walls are flat.
Oblate Spheroidal Waveguide:
Unity Horn
An externally hosted image should be here but it was not working when we last tested it.
The Peavey Quadratic Throat Waveguide would be closer to ideal if it was round, but it's still a waveguide
http://aa.peavey.com/downloads/pdf/qwp1.pdf
The Bastanis Prometheus uses a VERY short throat, but that is indeed a waveguide
Audio Kinesis
An externally hosted image should be here but it was not working when we last tested it.
JBL M-Pro
Kudos to Dan - I think he found the most affordable waveguide in the world! JBL part #354267-001.
There are two exceptions to the rule that the walls are flat. At the throat and at the mouth the flat walls of the waveguide must mate to the throat and the enclosure, respectively. You'll notice that most of the waveguides above change angle at the throat and the mouth for this reason. The change in angle should be as gradual as possible, to reduce HOMs.
doug20 said:
For a decade I was involved in the tuner scene, fixing up cars. I tried different rims, tires, tweaks to the power train, suspension, etc... I knew a bunch of people who were involved in the scene also, and we would all daydream about what we would do to our cars next. Our cars were always a work in progress.
Recently I had the opportunity to drive a real sports car. And it was a REVELATION. I just couldn't believe how much fun it was to drive. I'd been fixing up cars for almost a decade and a half, and I'd never even bothered to drive a real sports car. Honestly, I felt like I'd wasted a decade screwing with cars, when I should have just BOUGHT a sports car. If I'd taken all the money I'd poured into tweaks and just bought a used sports car, the results would have exceeded anything I could do on my own.
A car is more than the sum of it's parts. You can add a turbo and big rims to a Honda Civic, but it's still not going to be a Porsche 911. As soon as you settle into the deeply sculpted seats, smell the leather, turn the key, and hear the motor rumble, you can tell this is a real sports car. Words cannot describe the difference in the experience, you just have to drive one.
I feel the same way about waveguides versus horns. We can talk about directivity, distortion and frequency response until we're blue in the face. But it's all academic until you actually sit down and LISTEN to one.
gedlee said:
The JBL product would be the JBL waveguide that is primary to this thread. Its what the thread started on and I as just trying to get back to what the measurements show for that JBL waveguide.
Patrick Bateman said:
First thanks for all the pics, lots of options. Maybe I should just buy an old JBL MP415 or something.
I could just spend $30K and be done with it, I could have done that when I finished my house but Im not searching for the ultimate sound so sitting down and listening isn't going to matter.
Im just interesting in the differences....not if they are crap, great, okay...just the differences and I want to learn how to spot those differences through measurements and by buying as much stuff to test as possible.
"If 90 x 40 could be done with both axes ideal then yes, this would be perfect. But such a thing is not possible."
What would happen if arrayed two 45 deg WG's next to each other?
Do the combined wavefronts destroy the horizontal CD?
What would happen if arrayed two 45 deg WG's next to each other?
Do the combined wavefronts destroy the horizontal CD?
doug20 said:
I have a kind hint on other work than Geddes' regarding horn ./. waveguide & sound perception in audio:
http://134.130.184.8/opus/frontdoor.php?source_opus=8
citation:
Listening tests were conducted to further investigate why different loudspeakers still sound different, even when all equalised to the same on-axis frequency responses. It is shown that these speakers still produce quite different ear signals (as may be demonstrated by dummy head measurements), probably caused by different radiation properties and presumably also by different wave front patterns. Once different loudspeakers are equalised to produce the same in-ear frequency response, they become virtually indistinguishable. Hence, phase errors and non-linear distortion seem to have only negligible effect on the perceived differences between loudspeakers.
The last sentence addresses "HOM" - phase errors that is reflection from diffraction with pathlengths and interference - and harmonics. HOM is no issue. Good news, really! The referred paper is a dissertation (for a phd). Hence it is of approved scientific value. The separation of horns/waveguides by having no HOM (Geddes) and having HOM (any other) may have no substance in the end.
so long
doug20 said:
IMHO, the magic is in the waveguide. If you get the WG wrong, no amount of time or money is going to fix it. But buying the right waveguide goes a long way towards getting "the ultimate sound."
That's the whole reason I noted the JBL M-PRO a couple of days ago; it's a really nice waveguide, and no one is using it. Kudos to Dan for finding another one.
Every time I've tried to remove the harshness from a horn via EQ, I've failed. I've listened to horns that measure flat, but still have that raspy coloration in the upper midrange.
In my experience, a cheap compression driver on a good waveguide sounds better than an expensive compression driver on a good horn.
I hope this thread wins a few people over to The New School 😀
Patrick Bateman said:
If by "new School" you are talking about Geddes then I would say that has already happened. Horn/waveguide discussion are hot button topics and everyone is buying pro audio parts (JBL, QSC ,etc) to test theories.
Even though he thinks I insult him I have posted many times that Im following his theories, etc.
That JBL waveguide dan posted is close to the JBL 415s?
gedlee said:
It still works with any entrance angle. Say for a 90x50, calculate two OS curves, one for a 90 degree waveguide and one for a 50 degree waveguide, both starting with whatever initial radius and angle you need. You'll end up with an r90 and and r50 for each slice z. At z=0, r90 = r50. Then, for each z as you move out, generate xy coordinates for the ellipse using the formulas x = r90*cos(theta) and y = r50*sin(theta) where theta is the radial angle around the ellipse.
xpert said:
It is funny to me that the very first sentence tells me exactly why having a proper CD waveguide is so important. Equalizing the on-axis frequency response and not being ultra concerned about the overall polar response will produce the results noted in the study. Why wouldn't you want to be concerned about the radiateed energy and it's response characteristics?
You are certainly onto something here ex-pert. Keep up the great work.
doug20 said:
You feel my pain, obviously. 😉
I could NOT get the issues discussed, as may be seen by reading the initial 20 pages of this thread, in which most everyone pursued their private off-topic agendas. Earl's purpose was clear from the beginning: dismiss this cheap alternative. Look back to the predecessor thread which initiated this. NO WAY could it hold +/- 2 dB out to 30°. Instead, he showed that it did, and better, in fact, than his standard, ESP12.
We ran with it, the rest is history, and now, a mere year later, literally hundreds of DIYers are exploring and enjoying the pleasures of the fundamental principles Earl teaches with minimal investment. If that's a challenge to anyone's perceptions, I'll not be apologizing for any of it. Ever.
doug20 said:
While there's a great deal of interest in waveguides, the majority of the implementations are faulty in one aspect or another. I think that's how I annoyed a bunch of people over at Avsforum. People really don't like to hear that their shiny new project is "compromised."
But engineering always requires compromises.
Maybe I should come up with a nicer way of saying that?
Here's a few pics of what I mean by "compromises."
Exhibit A: The JBL Progressive Transition Waveguide:
In the side view, you can see there's a gentle curvature of the walls. This curvature will narrow the directivity at high frequencies. It's almost a waveguide, but not quite. If the walls were flat, and the termination at the mouth was smoother, it would be a waveguide. Take a look at the curves I posted here, you can see the difference in the graphs. http://www.diyaudio.com/forums/showthread.php?postid=1838158#post1838158
Thanks to Zilch for the pic, which I stole from Lansing Heritage.
Exhibit B: The 18Sound XT1086
These guys got soooooo close. This stupid waveguide is almost perfect, it's frustrating that they dropped the ball. It has constant coverage, the mouth termination is impeccable, it's sturdy, and it's affordable.
So why did they screw up the throat???
I took a picture of the one that I bought, then I cranked up the gamma so that you can see what's wrong with it. It's difficult to photograph, because it's black. I looked all over the internet for a pic of the throat, and couldn't find one. But the throat is totally screwed up. Check out the pic; see how it's elliptical at the throat? In addition, there's an abrupt change in angle, which will cause gobs of HOMs. It's a shame really, this is such a good design in all other aspects.
The waveguide on the left is the MCM, and the mold on the right is my own design, for my car. It's triangular to mate with a car's windshield.
Exhibit C:
All those waveguides with dome tweeters
When you use a dome tweeter, the directivity falls apart at high frequencies. To make matter worse, most of the waveguides for dome tweeters only work at high frequencies LOL
That's because they're so shallow.
So these designs give you some gain, and they make it easier to crossover by aligning the voice coil of the drivers. But they're compromised. (There I go with that word again...)
An externally hosted image should be here but it was not working when we last tested it.
Of course engineering always requires compromises, so it's no surprise that the most popular "waveguides" are compromised. Going back to my car analogy, the Mazda Miata outsells the Porsche 911 by a huge margin because it's a great sports car that's affordable. It might not be as uncompromising as a 911, but it's still a great sports car.
Maybe we can all work together and find the audio equivalent of a Mazda Miata; a great design that everyone can afford.
FWIW, John Janowitz took some good pictures of the XT1086 that (I think) show what you're talking about:
noah katz said:
This may be possible, but it would not be my first choice. An elliptical waveguide appears to be a good option once you get the details worked out.
xpert said:
Spin doctoring all the time, really.
It is only wishful thinking on your side that this is adressing HOMs. In fact the author may not even have heard of HOMs.
The original German paper may be of approved scientific value. The translation definitely is NOT. The German word "untergeordnet" can´t be translated as "neclectible". The correct translation would be "secondary" in this case, which is quite different from "neclectible". No one would argue that frequency response comes first and nonlinear distortion comes second.
Need I mention that nothing like "virtually indistinguishable" is said in the original German abstract?
catapult said:
That is a substantial modification to your initial post, but is quite astute. I still think that you might run into some glitches, as I did, but that path is basically the same one that I followed to get my mold made.
Saurav said:
Ah nice! Much better than the ones I took.
It's such a great waveguide, in most respects.
Constant directivity? Check.
Gentle termination to the baffle? Check.
Elliptical mouth, to eliminate the dip? Check.
Not too elliptical, so the power response is decent? Check.
Not too expensive? Check.
Not too shallow? Check.
It's not back-ordered for two months? Check.
I'm half tempted to file down the throat and do some polars on it...
ZilchLab said:
You know that really is rude. I did this study objectively and presented the data accurately. That fact that you see what you want to and that's NOT what I see does not in any way make my analysis any less objective. If anyone here has prejudices it's you!!
xpert said:
If you don't care that your speakers sound a little like a megaphone, then no HOM is no issue. Don't worry, be happy.
But I don't like it, and it's clearly audible. It doesn't take special golden ears. It's more audible or, to me objectionable, with some music than others. Depends on the frequencies affected. In addition to the Unity's, I made a pair of small back loaded horns, which clearly have horn character, but at lower frequencies than the waveguides in this thread. Same thing though. Sounds like the music played through a larger horn, that's all. I can listen to them for background music, but not for when I want to concentrate on the music. BTW, if I stuff them more, the megaphone character goes away, but then so does the bass. Kinda defeats the purpose.
Here's a test a xpert or novice can perform, if they have any interest at all in the actual topic: http://www.audioheritage.org/vbulletin/showthread.php?t=12967 If you hear no difference, then you can have all the dynamics of a horn and not worry much about the profile, making this thread irrelevant for you. One less thing to worry about.
Sheldon
Patrick Bateman said:
It would be easy to make a low cost "LOW compromise" waveguide as an injection molded part. Just write me a check for the $60,000-80,000 tooling and I'll sell them at cost plus 20% for handling. They would be about $15 each.
Now I personally don't like ABS molding at all because its not very strong or well damped. And I don't like the driver mounting where the thinest section is right were the most weight is. And fixing these things jacks up the price quite a bit. And then there is the foam plug, well those aren't cheap any way you "cut it" (pun intended).
The point is that my target market never has been a "low cost" solution because THERE IS NO MARKET there. When one is willing to compromise on cost then there is basically no end to the compromises that start to happen and you basically end up with exactly what you see in the market. A "one size fits all" compromise solution that IS low cost and IS a serious compromise.
If people like the JBL horn then fine, use it. Its not for me thanks. I listened to JBL horns for almost 30 years. Thats why I looked for another solution.
- Status
- Not open for further replies.