The Earthworks is very good all the way out to 10 kHz and yes, its very small. Doesn't have to be big. My only point is that mics are darn near perfect reproducers of sound. The low efficiency allows this. There is no room to improve microphone technology its totally mature.
The higher end mic I bought to replace the Behringer was a used but recently calibrated B&K 4007 mic. I was told it would likely fit my needs. I have no idea how it compares to the Earthworks, but I've heard good things about those too. They just don't come up used at good prices very often. I also have a 4149 coming that I picked up used on ebay for like 75 dollars. We will have to see its condition and accuracy though.
Most applications use parabolic dishes for directivity if that is what you are looking for.jzagaja said:Can OS waveguide be used in a reverse as a microphone or the requirements are completely different?
Panasonic WM61 capsules are quite smooth at the top end up to around 20KHz, that is why they are used often for measurements. But the noise level is higher than most people would like for high quality recordings.jzagaja said:But condenser mic maybe except very small ones usually doesn't have smooth top end like a good velocity mic? Again diffraction matters.
Search the Yahoo mic builders user group and you can find more information.
Dr. Geddes, could you elaborate on that? what types of mics? What happens as they age? How long does it take for them to become "not well"?gedlee said:That is the unfortunate thing about mics - they don't age well.
Its highly variable. The diaphragm tension is critical and this varies with age and humidity depending on the material. If its an electret then the charge leaks off. There is usually an FET in the capsule and this changes with time. There is a load resistor that can be in the gigaohms (we even had them in the TeraOhms) whose resistance can change. Everything is design dependent. making a microphone is easy. Making a calibrated microphone is tougher and making one that does not change over time is extremely difficult.
Let's get back to horns/waveguides for a moment.
First, I think I misread your vertical off-axis charts earlier. You have the top few lines similar in color and they're closely grouped so it is easy to misread them. I thought the nulls start in the upward direction at 15° but after looking closer, it appears they start forming around 22.5°. It appears the downward nulls start forming around 7.5°, forming an arc of about 30° where summation is constructive.
Second, you stated summing is coherent even when destructive and I agree with you. I misspoke. I chose my words poorly and should have said constructive instead of coherent.
However, I still think we have some disagreements and I'd like to look at those. You said a listener is never going to be at a null in the Summas and I think unless you give specific guidance on placement, that is too optimistic. If a person were to put the speaker on the stands as I've seen at shows and in photos, there is a very real possiblity they will be located in a null.
The off-axis null starts growing very rapidly as you move under the forward axis. By 7.5°, it is more than -6dB at 1.5kHz and over a fairly broad range. At larger off-axis angles, the null deepens but it is already worse than -6dB at 7.5°. This is a pretty shallow angle.
If you considered the forward axis to be the centerline of the speaker, then 7.5° down is basically straight on-axis with the woofer for several feet back. Even at larger distances, it isn't far off the forward axis. So I think it is important to let people know this, and not brush it under the rug saying "horizontal is everything, vertical is only somewhat important". Horizontal is more important, but not if you're sitting in a vertical null.
Another thing I want to address is the description of off-axis response in the vertical plane. I've described it in the past with illustrations and you've said I was "arm waving" and other such dismissive phrases.
You've said "show me the data", sometimes right after a post I've made that showed you the data. So I've decided that the best way to talk about this with you is to use data you've gathered yourself.
The measurements you made comparing the Yamaha speaker with your Summa illustrate it so well, I've used your measurements to describe the benefits of reducing the horn's vertical angle. The Yamaha speaker doesn't have very good on-axis curves, and its polars aren't any better. But the general trend still shows very well.
It has always been my position that matching directivity in the horizontal was important, and that the vertical angle should be limited to the null angle as much as possible. Sometimes I think you agree with this, but most times you don't, sometimes seeming to dismiss the vertical pattern entirely.
In this thread, you've admitted there is benefit in making the drivers closer together to increase the null angle. You've admitted there is benefit in reducing vertical pattern to a narrow angle. What I don't understand, is what exactly do you disagree with? Would you prefer a more narrow vertical pattern or not?
I think I understand that you have not had a good way to build such a horn. If I remember right, you have said it is mathematically possible but fabrication of the horn was more difficult. That makes sense to me. If you can't make such a thing, then obviously it isn't an option for you. But I seem to recall that you said you had found a way to fabricate a horn recently.
What I'm saying is, if you can make such a horn, wouldn't that be better? If you can make a horn with good horizontal coverage over a 90° arc and vertical pattern limited to 40° or so, wouldn't that be attractive to you? If you can make it so the sound sources can be spaced closer together, widening the null angle, wouldn't that be better? The mouth will be less tall, so it will lose some vertical control down low, but with interference nulls down there anyway, don't you consider this a useful trade-off?
First, I think I misread your vertical off-axis charts earlier. You have the top few lines similar in color and they're closely grouped so it is easy to misread them. I thought the nulls start in the upward direction at 15° but after looking closer, it appears they start forming around 22.5°. It appears the downward nulls start forming around 7.5°, forming an arc of about 30° where summation is constructive.
Second, you stated summing is coherent even when destructive and I agree with you. I misspoke. I chose my words poorly and should have said constructive instead of coherent.
However, I still think we have some disagreements and I'd like to look at those. You said a listener is never going to be at a null in the Summas and I think unless you give specific guidance on placement, that is too optimistic. If a person were to put the speaker on the stands as I've seen at shows and in photos, there is a very real possiblity they will be located in a null.
The off-axis null starts growing very rapidly as you move under the forward axis. By 7.5°, it is more than -6dB at 1.5kHz and over a fairly broad range. At larger off-axis angles, the null deepens but it is already worse than -6dB at 7.5°. This is a pretty shallow angle.
If you considered the forward axis to be the centerline of the speaker, then 7.5° down is basically straight on-axis with the woofer for several feet back. Even at larger distances, it isn't far off the forward axis. So I think it is important to let people know this, and not brush it under the rug saying "horizontal is everything, vertical is only somewhat important". Horizontal is more important, but not if you're sitting in a vertical null.

Another thing I want to address is the description of off-axis response in the vertical plane. I've described it in the past with illustrations and you've said I was "arm waving" and other such dismissive phrases.
You've said "show me the data", sometimes right after a post I've made that showed you the data. So I've decided that the best way to talk about this with you is to use data you've gathered yourself.
The measurements you made comparing the Yamaha speaker with your Summa illustrate it so well, I've used your measurements to describe the benefits of reducing the horn's vertical angle. The Yamaha speaker doesn't have very good on-axis curves, and its polars aren't any better. But the general trend still shows very well.
It has always been my position that matching directivity in the horizontal was important, and that the vertical angle should be limited to the null angle as much as possible. Sometimes I think you agree with this, but most times you don't, sometimes seeming to dismiss the vertical pattern entirely.
In this thread, you've admitted there is benefit in making the drivers closer together to increase the null angle. You've admitted there is benefit in reducing vertical pattern to a narrow angle. What I don't understand, is what exactly do you disagree with? Would you prefer a more narrow vertical pattern or not?
I think I understand that you have not had a good way to build such a horn. If I remember right, you have said it is mathematically possible but fabrication of the horn was more difficult. That makes sense to me. If you can't make such a thing, then obviously it isn't an option for you. But I seem to recall that you said you had found a way to fabricate a horn recently.
What I'm saying is, if you can make such a horn, wouldn't that be better? If you can make a horn with good horizontal coverage over a 90° arc and vertical pattern limited to 40° or so, wouldn't that be attractive to you? If you can make it so the sound sources can be spaced closer together, widening the null angle, wouldn't that be better? The mouth will be less tall, so it will lose some vertical control down low, but with interference nulls down there anyway, don't you consider this a useful trade-off?
As usual Wayne, you have it wrong. The axis for the polars is relative to the center of the system NOT the center of the waveguide. So 7.5 ° at 12 feet, as in my room, is at the bottom edge of the cabinet. No one is ever going to be that low unless they are laying on the floor.
The main point that I want to make is that I am not going to argue with you. I thought that maybe it was I who had a problem with you, but then I find that so does Tom Danley, and for the exact same reason. If you choose to treat Tom and I with the kind of disrespect that you have shown, then I don't feel obliged to discuss matters with you.
And no, its not "feasible" for me to make an elliptical waveguide at this time. I've explained this before, go back and read it.
The main point that I want to make is that I am not going to argue with you. I thought that maybe it was I who had a problem with you, but then I find that so does Tom Danley, and for the exact same reason. If you choose to treat Tom and I with the kind of disrespect that you have shown, then I don't feel obliged to discuss matters with you.
And no, its not "feasible" for me to make an elliptical waveguide at this time. I've explained this before, go back and read it.
Wayne Parham said:If you considered the forward axis to be the centerline of the speaker, then 7.5° down is basically straight on-axis with the woofer for several feet back...
Then you calculated wrong, because its 18" at 12 ft which is at the bottom of the enclosure as I said.
It's an easy calculation to make, too easy to get wrong.
You're the one that mentioned 12 feet. I said "several" feet.
When I wrote that comment, I was visualizing making measurements at 2 meters, basically six feet. At that point, the null is pretty much dead on with the woofer. So a listener 6 feet away, same height as the woofer might think they're pretty much on-axis, but is actually sitting in a null.
Even sitting 12 feet back, having a null at the bottom edge of the cabinet is too close, in my opinion. How many people put Summas on stands without knowing how close their listening position is to a vertical null?
The null angle is just too close to the forward axis, in my opinion. I would prefer it were further outside. I would design the speaker for the nulls to be further outside. If the listener is in a null, not much else matters.
I'm sorry if you find these statements to be offensive or rude. I do not mean them to be. They are just facts.
In your defense, you can always tell people to position your speakers with bias towards the area over the speaker. It is better on axis and above. That's a reasonable solution. But they still have a pretty narrow arc, only about 30° overall.
My point is, and always has been, that the vertical pattern is important because vertical nulls can be severe. If the vertical radiation angle is limited at HF and the nulls are placed right, they can mark the edge of the pattern. Seems the best way to do it.
You're the one that mentioned 12 feet. I said "several" feet.
When I wrote that comment, I was visualizing making measurements at 2 meters, basically six feet. At that point, the null is pretty much dead on with the woofer. So a listener 6 feet away, same height as the woofer might think they're pretty much on-axis, but is actually sitting in a null.
Even sitting 12 feet back, having a null at the bottom edge of the cabinet is too close, in my opinion. How many people put Summas on stands without knowing how close their listening position is to a vertical null?
The null angle is just too close to the forward axis, in my opinion. I would prefer it were further outside. I would design the speaker for the nulls to be further outside. If the listener is in a null, not much else matters.
I'm sorry if you find these statements to be offensive or rude. I do not mean them to be. They are just facts.
In your defense, you can always tell people to position your speakers with bias towards the area over the speaker. It is better on axis and above. That's a reasonable solution. But they still have a pretty narrow arc, only about 30° overall.
My point is, and always has been, that the vertical pattern is important because vertical nulls can be severe. If the vertical radiation angle is limited at HF and the nulls are placed right, they can mark the edge of the pattern. Seems the best way to do it.
"Almost 19""
But not 8" as Wayne claimed. That's exactly the problem, things are exagerated and hammered on until he thinks that he has won.
Wayne has some valid points, like IF I could readily make an elliptical waveguide that I should try it. That is certainly true and I'd love to do that, but, as I have said over and over again, I can't. So whats the point of beating it to death? Some other agenda maybe?
But not 8" as Wayne claimed. That's exactly the problem, things are exagerated and hammered on until he thinks that he has won.
Wayne has some valid points, like IF I could readily make an elliptical waveguide that I should try it. That is certainly true and I'd love to do that, but, as I have said over and over again, I can't. So whats the point of beating it to death? Some other agenda maybe?
For the record, my issue with Wayne does not stem from what he says here, it's something that happened several years ago.
Wayne asked me to speak at his GPAF, and I agreed. When he found out that I was going to speak on the subject of multiple subs in a small room, he went ballistic because he did not agree with my recommendations for random placement (which I have proven over and over again, including a proof posted on my web site). To make a long story short, he told me, in a very rude manner, that unless he could approve of what I was going to say to not come at all. As a scientist that kind of censureship and rudeness is unacceptable. So I didn't go.
Wayne asked me to speak at his GPAF, and I agreed. When he found out that I was going to speak on the subject of multiple subs in a small room, he went ballistic because he did not agree with my recommendations for random placement (which I have proven over and over again, including a proof posted on my web site). To make a long story short, he told me, in a very rude manner, that unless he could approve of what I was going to say to not come at all. As a scientist that kind of censureship and rudeness is unacceptable. So I didn't go.
horn format
Wayne
i do not have much understanding of the theorys, you expose here, but had already a number of different horn speakers, and different kind of horns : radial horns, rectangular horns like you use, sectorial multi cellular horns, and tractrix horns. The far worst and most colored sounding horns where the rectangular ones you use. They screw up sound entirely, honk and sound horny, and i would never use them again in a home setup. They might be o.k. for P.A. but for hi-fi ? never again. Interestingly, the multi cell horns behave quit different than just regular rectangular horns. Why - i don' t know.
Wayne
i do not have much understanding of the theorys, you expose here, but had already a number of different horn speakers, and different kind of horns : radial horns, rectangular horns like you use, sectorial multi cellular horns, and tractrix horns. The far worst and most colored sounding horns where the rectangular ones you use. They screw up sound entirely, honk and sound horny, and i would never use them again in a home setup. They might be o.k. for P.A. but for hi-fi ? never again. Interestingly, the multi cell horns behave quit different than just regular rectangular horns. Why - i don' t know.
FWIW, an idea on making the ellipse:
Make a round wave WG with with walls thin enough to be flexible.
Make a 2-piece elliptical front flange that clamps around the mouth and deforms the WG into the ellipse.
This flange would have a proper large radius and a rebate to capture the front of the WG.
If it's Murphy's day off, the shape of the transition from round to elliptical will be close enough to what it ought to be.
Smear whatever on the outside walls of the WG to stiffen/damp them.
Make a round wave WG with with walls thin enough to be flexible.
Make a 2-piece elliptical front flange that clamps around the mouth and deforms the WG into the ellipse.
This flange would have a proper large radius and a rebate to capture the front of the WG.
If it's Murphy's day off, the shape of the transition from round to elliptical will be close enough to what it ought to be.
Smear whatever on the outside walls of the WG to stiffen/damp them.
noah katz said:
If it's Murphy's day off, the shape of the transition from round to elliptical will be close enough to what it ought to be.
Murphy's at work today The contour would have to be wrong.
The sole issue with the elliptical is tooling cost - it sky-rockets. And this cost has to be passed on to the product. And what if it doesn't work any better? Then what?
Its not like people complain or that I have ever heard a problem caused by the nulls at the floor. In a seated listening position it is simply NOT an issue. No matter how hard one tries to make it one.
I'll have to warn my wife that the sound won't be as good the next time we "do it" on the floor in the HT. We'll have to move to the couch I guess.
"The contour would have to be wrong."
I thought the goal is minimum rate of change of curvature moving down the WG axis, and that this may be exactly the result of mechanically deforming the WG as the walls assume the least stress condition.
If not, nonlinear FEA could be used to determine what variation in wall thickness would give the desired end result, though this too could get expensive.
Or just carry the deformation concept further and make a series of 2-piece rectangles with elliptical cutouts spaced along the WG's length.
They could be MDF, made with CNC or manually with patterns.
BTW I agree on the vertical null issue, as I don't see a scenario where the listener would be that low and close to the speakers.
I thought the goal is minimum rate of change of curvature moving down the WG axis, and that this may be exactly the result of mechanically deforming the WG as the walls assume the least stress condition.
If not, nonlinear FEA could be used to determine what variation in wall thickness would give the desired end result, though this too could get expensive.
Or just carry the deformation concept further and make a series of 2-piece rectangles with elliptical cutouts spaced along the WG's length.
They could be MDF, made with CNC or manually with patterns.
BTW I agree on the vertical null issue, as I don't see a scenario where the listener would be that low and close to the speakers.
I know exactly "how to" make the elliptical waveguide, I just don't see the urgency in risking that kind of time and expense. When the product needs updating then I'll look into doing it, but for now I have my hands full with prototype development and orders.
- Home
- Loudspeakers
- Multi-Way
- Geddes on Waveguides