Re: Re: Re: Re: Where to get the QSC?
Honestly, I'm not sure it makes any sense at all to talk about performance at oblique angles when the vertical null angle is as small as a round horn necessitates (because of driver spacing). I mean, it seems like you'd want to take care of the verticals before worrying about the obliques. I think probably once the round horn guys move over to the ellipticals, then they'll start getting somewhere but for now, their verticals and obliques are kind of a mute point, if you ask me.
Round horn / waveguide guys sometimes talk about their horns as if they weren't used in conjunction with other drivers. If that were the case, round horns would be great. But they're not, so it doesn't really matter what the vertical pattern control looks like, and for that matter, not much use talking about diagonals. The forward lobe is squished pretty thin, and the upper and lower side lobes are much larger by comparison.
Earl Geddes at gedlee.com[/i] The first two plots are the horizontal polars. The next two are the vertical polars and the last two are an an oblique angle of about 45°. Rectangular horns often have poor response along this oblique angle.[/QUOTE] [QUOTE][i]Originally posted by ZilchLab said:
Honestly, I'm not sure it makes any sense at all to talk about performance at oblique angles when the vertical null angle is as small as a round horn necessitates (because of driver spacing). I mean, it seems like you'd want to take care of the verticals before worrying about the obliques. I think probably once the round horn guys move over to the ellipticals, then they'll start getting somewhere but for now, their verticals and obliques are kind of a mute point, if you ask me.
Round horn / waveguide guys sometimes talk about their horns as if they weren't used in conjunction with other drivers. If that were the case, round horns would be great. But they're not, so it doesn't really matter what the vertical pattern control looks like, and for that matter, not much use talking about diagonals. The forward lobe is squished pretty thin, and the upper and lower side lobes are much larger by comparison.
Re: Re: Re: Re: Re: Where to get the QSC?
Earl has promised to demonstrate that the vertical nulls and lobes are not an issue if we merely listen off-axis as he prescribes.
[There's clearly no point in showing the on-axis verticals with TWO holes in the response....
]
Wayne Parham said:
Earl has promised to demonstrate that the vertical nulls and lobes are not an issue if we merely listen off-axis as he prescribes.
[There's clearly no point in showing the on-axis verticals with TWO holes in the response....
]noah katz said:
Yes, sorry if that was not know. my registration for penngray wasnt working (have no idea why) so I used doug.
Vertical nulls
Well, yeah, I know the usual response from that camp. Ironically, it's the same thing the tractrix guys say. They don't care what happens outside the forward axis. Ten years ago, the tractrix round horns were all the rage, and the arguments from tractrix supporters sound very much the same as the ones used by the round OS horn guys today.
You know, if the difference were only a few degrees, I wouldn't see a problem. Say if the rectangular / elliptical could only push the nulls out five degrees compared to the round horns, then I think the round horns / waveguides would be more of a viable option. If, say the speaker with a rectangular horn produced a 50° arc between vertical nulls and the same speaker with a round horn made a 40° pattern, then I'd find the trade-off leaning less towards the rectangular or elliptical horn. But when you're looking at patterns that are closer to half that, and particularly if (when) one of the nulls is shifted up close to the surface normal of the baffle, I think that's a real problem. That's what makes asymmetrical horns so much more attractive to me.
I was particularly frustrated to see a few guys try to obfuscate this fact, even going so far as to poke fun at the idea of caring about the position of the vertical nulls. In my opinion, these kinds of statements were an effort to minimize the problem, really a form of deception. I pretty much bowed out at that point, but it was frustrating for me to see what I thought was irresponsible, even deceptive.
I think you probably remember that I pointed out the Summa had a vertical null only 7.5° down from the baffle surface normal. I probably wouldn't have pointed that out with an unsolicited comment, but it was in response to a comparison made to speakers that I think are frankly, superior. When you calculate the position of the 7.5° null, you see that it is straight in front of the speaker for over ten feet back. Basically, you're sitting squarely in a null if you're facing the speaker within ten feet from it. There's a huge notch in response. So to me, that's a pretty big deal.
The reply from some of these guys was that they didn't listen to the speaker from the floor. OK, well, do they put the speaker on stands? Because that's how I saw it displayed. I mean, what's with the attitude? The fact is, the null is straight in front of the speaker, measurements show it, and it isn't right to minimize that fact by joking that you don't listen on the floor. It isn't at floor level until you're at least 10 feet back, and if the speakers are on stands, even further back than that.
Why that was frustrating to me is that I'm basically in the same camp. We think a lot alike. In fact, I'd consider most of the round horn guys to be personal friends. We're in agreement in most respects. I like smoothly radiused horns that provide uniform directivity. The catenary profile appears to be the best shape. I think all of us interested in this thread want a uniform reverberent field. We all want limited diffraction. We all want good sound and most of the things it takes to get good sound we agree upon. But for some reason, the round horn guys don't want to look at the vertical nulls.
As for me, I say fix the forward lobe first, then reduce diffraction within the constraints that keep the forward lobe right. Kinda doesn't matter much how low the HOM if you have a 10dB notch in response straight out in front of the speaker. Not only does the round horn have the narrow vertical null angle, but it has that dip on-axis because of mouth edge diffraction. It just seems to me there are a lot of benefits in going asymmetrical.
ZilchLab said:Earl has promised to demonstrate that the vertical nulls and lobes are not an issue if we merely listen off-axis as he prescribes.
[There's clearly no point in showing the on-axis verticals with TWO holes in the response....
Well, yeah, I know the usual response from that camp. Ironically, it's the same thing the tractrix guys say. They don't care what happens outside the forward axis. Ten years ago, the tractrix round horns were all the rage, and the arguments from tractrix supporters sound very much the same as the ones used by the round OS horn guys today.
You know, if the difference were only a few degrees, I wouldn't see a problem. Say if the rectangular / elliptical could only push the nulls out five degrees compared to the round horns, then I think the round horns / waveguides would be more of a viable option. If, say the speaker with a rectangular horn produced a 50° arc between vertical nulls and the same speaker with a round horn made a 40° pattern, then I'd find the trade-off leaning less towards the rectangular or elliptical horn. But when you're looking at patterns that are closer to half that, and particularly if (when) one of the nulls is shifted up close to the surface normal of the baffle, I think that's a real problem. That's what makes asymmetrical horns so much more attractive to me.
I was particularly frustrated to see a few guys try to obfuscate this fact, even going so far as to poke fun at the idea of caring about the position of the vertical nulls. In my opinion, these kinds of statements were an effort to minimize the problem, really a form of deception. I pretty much bowed out at that point, but it was frustrating for me to see what I thought was irresponsible, even deceptive.
I think you probably remember that I pointed out the Summa had a vertical null only 7.5° down from the baffle surface normal. I probably wouldn't have pointed that out with an unsolicited comment, but it was in response to a comparison made to speakers that I think are frankly, superior. When you calculate the position of the 7.5° null, you see that it is straight in front of the speaker for over ten feet back. Basically, you're sitting squarely in a null if you're facing the speaker within ten feet from it. There's a huge notch in response. So to me, that's a pretty big deal.
The reply from some of these guys was that they didn't listen to the speaker from the floor. OK, well, do they put the speaker on stands? Because that's how I saw it displayed. I mean, what's with the attitude? The fact is, the null is straight in front of the speaker, measurements show it, and it isn't right to minimize that fact by joking that you don't listen on the floor. It isn't at floor level until you're at least 10 feet back, and if the speakers are on stands, even further back than that.
Why that was frustrating to me is that I'm basically in the same camp. We think a lot alike. In fact, I'd consider most of the round horn guys to be personal friends. We're in agreement in most respects. I like smoothly radiused horns that provide uniform directivity. The catenary profile appears to be the best shape. I think all of us interested in this thread want a uniform reverberent field. We all want limited diffraction. We all want good sound and most of the things it takes to get good sound we agree upon. But for some reason, the round horn guys don't want to look at the vertical nulls.
As for me, I say fix the forward lobe first, then reduce diffraction within the constraints that keep the forward lobe right. Kinda doesn't matter much how low the HOM if you have a 10dB notch in response straight out in front of the speaker. Not only does the round horn have the narrow vertical null angle, but it has that dip on-axis because of mouth edge diffraction. It just seems to me there are a lot of benefits in going asymmetrical.
Re: Vertical nulls
And those would be yours of course. No surprise there. Your numbers are still incorrect by the way.
Wayne Parham said:
And those would be yours of course. No surprise there. Your numbers are still incorrect by the way.
Re: Re: Vertical nulls
The "numbers" may be unflattering, but they are correct.
They even came from your measurements.
gedlee said:
The "numbers" may be unflattering, but they are correct.
They even came from your measurements.
Member
Joined 2003
Wayne,
For the same relative LF pattern control, would the mouth of a 90*x45* asymmetric horn be twice as wide as a 90* axisymmetric? If not, what is the correct relationship (in wavelengths or physical size)?
Thanks,
Paul
For the same relative LF pattern control, would the mouth of a 90*x45* asymmetric horn be twice as wide as a 90* axisymmetric? If not, what is the correct relationship (in wavelengths or physical size)?
Thanks,
Paul
Paul W said:
It actually depends on the flare profile, but I assume you're not talking about any of the diffraction horns. For waveguides and horns like radials and ellipticals, the aspect ratio roughly matches the pattern.
Member
Joined 2003
Yes, I'm asking about devices with minimal diffraction...but I don't think I phrased my question clearly enough.
If the mouth of a 90 degree axisymmetric is 10" high, would a 90x45 degree asymmetric with the same frequency coverage and degree of pattern control be 10" high by 20" wide?
If the mouth of a 90 degree axisymmetric is 10" high, would a 90x45 degree asymmetric with the same frequency coverage and degree of pattern control be 10" high by 20" wide?
Here's a pic of the lobing of a 'Summa-like' speaker with an 800 Hz crossover and 16" c-c spacing. The nulls are separated by about 60 degrees. Now the actual tilt of the lobe would be determined by specifics of the crossover and the acoustic centers of the drivers but there's certainly nothing about the 'big round horn' that stops you from having decent lobing performance if you cross low enough.
Attachments
And here's the same thing, decreasing the c-c distance to 13" like you might do with an elliptical or rectangular horn. The nulls move farther apart as Wayne says; now they're about 80 degrees apart. Whether or not 60 degrees vs. 80 degrees is an important tradeoff is left as an exercise for the viewer. 😉
Attachments
Re: Vertical nulls
That's my problem with it, as well. It's short-sighted, at best, for the proponents, the prime one, in particular, to so relentlessly defend the exclusive "one-way" posture they assume.
Earl's purpose for this thread was to categorically dismiss an economical alternative incorporating many of his important teachings, and his motivation was perfectly obvious from the outset, and well, "small." In retrospect, it may well have served more to illuminate the shortcomings of his own approach, instead.
With credibility comes an obligation to exercise it responsibly, in my view.... 🙁
Wayne Parham said:
That's my problem with it, as well. It's short-sighted, at best, for the proponents, the prime one, in particular, to so relentlessly defend the exclusive "one-way" posture they assume.
Earl's purpose for this thread was to categorically dismiss an economical alternative incorporating many of his important teachings, and his motivation was perfectly obvious from the outset, and well, "small." In retrospect, it may well have served more to illuminate the shortcomings of his own approach, instead.
With credibility comes an obligation to exercise it responsibly, in my view.... 🙁
catapult said:
The problem is the crossover frequency is largely set by the horizontal directivity. You want to set it where the direct radiating midwoofer and the tweeter match in the horizontal plane. Crossover at 800Hz is too low for proper matching of even a 15" midwoofer. The smaller the driver, the higher the frequency you'll want to crossover. So you have some competing priorities here.
Wayne, if you'd care to tell us the XO frequency and c-c spacing of your best 15" system, I'd be happy to run a lobing sim.
catapult said:
With a notch in the 30° vertical off-axis response at 800 Hz, just about right for corrupting the ceiling bounce.
catapult said:
That works well for the 12" Abbey, actually, if only we could cross it that low.
catapult said:
Just right to "punctuate" the narrower 50° vertical beamwidth of our axi-asymmetric elliptical or rectangular waveguide....
Help me, now, why is the lightest blue, higher frequency lobe wider in each instance; am I suffering a brain freeze here...?
catapult said:
You don't need a sim, here are measurements:
This isn't my best loudspeaker, but it is one that is a DI matched two-way, like those discussed here. You'll see nulls about 50 degrees apart.
A similar speaker with a round horn puts the nulls about 30 degrees apart.
- Geddes vertical measurements, comparing a Summa with Yamaha loudspeaker
One thing to consider when designing (and simulating, for that matter) is the position of the forward lobe depends not just on the crossover frequency and vertical spacing, but also the acoustic overlap between drivers, the crossover phase and the distance from listener to sound sources, set by their depth on the baffle. This sets the position of the forward lobe, the vertical nulls and the outer lobes. The size of the outer lobes is partially determined by the directivity of the individual sources too.
Real world loudspeakers rarely have the forward lobe centered on the baffle normal, with equally spaced vertical nulls and outer lobes. In fact, I personally have never seen one. The best speakers I've seen have the forward lobe within 5 degrees of the baffle normal, with nulls spaced fairly far apart, enough to be useful.
When working with physical components having set values (2.2uF, 3.3uF, etc) you do not have an unlimited number of choices to work with. You could get custom parts, I suppose, or you could hand pick values. But the point is that you're generally working with specific incremental values. You also have to work with the physical properties of the drivers and horns chosen. A horn or waveguide has depth that is largely determined by its flare profile and coverage pattern. The woofer and tweeter each have specific depth too. So the designer must juggle all these values, to find several solutions simultaneously. There are competing priorities to balance.
The point of all this is it is very unlikely that the forward lobe will match the baffle surface normal. Best case it's off by a few degrees, maybe five. In many cases, the forward lobe is shifted much further than that. When that happens, the vertical nulls will not necessarily fall above and below the speaker by the same amount. One may be very close to the baffle normal, which is what most people consider to be the central forward axis.
In my opinion, it is important that the forward lobe be close to centered and also that the nulls be fairly wide apart. Both things have to happen for the forward lobe to be clean and useful. This depends on crossover frequency and slope, acoustic overlap between drivers, and the physical relationships of the drivers.
ZilchLab said:
Because it's dominated by the tweeter so you don't see much woofer-tweeter interference with the LR4 crossover I was modeling. APRE doesn't do horns so I was faking horn directivity by using a large diameter tweeter. The location of the nulls is still accurate though.
- Status
- Not open for further replies.