The idea is to increase the usable listening area. A rather old idea (60's) by Ben Bauer. It works to some extend with every speaker that falls off off-axis.
The contralateral reflection is indeed stronger than the ipsilateral one. I found this to cause a somewhat narrower sound stage but it very much depends on how strong other reflections are.
My limited experience says that as long as the speakers cross in front of you the results are pretty similar. This would be true from the data as well since the speakers are pretty similar form 10° - 25° (i.e. 20° - 50° toe-in from normal).
Lets assume that we have a perfect 90° pattern. If the speakers are toed-in at 45° then the near wall is not illuminated at all, nor if the wall behind the speakers, but yes the far wall is illuminated. Anything less than this will illuminate the near wall. The near wall has both the earliest reflection as wll as the strongest one, and the reflection arrives dominately at the same ear as the direct sound. The far wall refelction is longer and lower in level and arrives dominately at the opposite ear. While there is no research to show that this is significant (its what Lidia and I are going to study) it seems clear that longer lower reflections to opposite ears (hence a great IACC) is by far the prefered scenario.
You can't think of the situation as "chance for reflections", you have to think about it as delay, level and direction. The delays should be as long as possible, the level as low as possible and the angle as close to pure lateral as possible. All of these things are optimized by a 90° coverage set at 45°.
Then there is the time-level stereo tradeoff benefit as one moves across the room allowing for a much large "sweet spot". This can only happen with toe-in as well.
Hence its a "strength" of waveguides that makes this possible.
I see. If the near wall is illuminated with less toe-in, then it makes sense. I was pretty sure it could be avoided in many room with something close to on-axis pointing.
In narrow rooms however, reflections from the opposite walls will arrive so early that it will be clearly audible and have a negative effect on accuracy. But less detrimental then the near wall. So yes, a better compromise if the sidewalls can't be treated. I'm quite sure this has been studied, but there are a lot of researches in certain acoustic communities that have never been published to the open.
Do your speakers measure as good if they are tuned to listen on-axis or would you have to make changes in the physical apperance as well?
In narrow rooms however, reflections from the opposite walls will arrive so early that it will be clearly audible and have a negative effect on accuracy. But less detrimental then the near wall. So yes, a better compromise if the sidewalls can't be treated. I'm quite sure this has been studied, but there are a lot of researches in certain acoustic communities that have never been published to the open.
Do your speakers measure as good if they are tuned to listen on-axis or would you have to make changes in the physical apperance as well?
Omholt
I think it best to not treat the side walls at all since this then gives some good lateral reflections in the reveberant field which are the key to a low IACC - i.e. good spaciousness. So its not a matter of "if the sidewalls can't be treated" - they shouldn't be treated.
The speakers measure well along any axis except directly on-axis. The waveguide has a very narrow hole right on axis because it is circular and the wavefront is so coherent (any other shape but circular or a less than ideal wavefront and this hole gets blurred.) If on-axis is desired then a crossover change can "modify" the hole, but then the power response at that frequency has a bump. The hole is fixable, but since I do not recommend on-axis listening it ends up being a non-issue. (In a three channel system the center channel has to be on-axis so the crossover gets modified.)
I think it best to not treat the side walls at all since this then gives some good lateral reflections in the reveberant field which are the key to a low IACC - i.e. good spaciousness. So its not a matter of "if the sidewalls can't be treated" - they shouldn't be treated.
The speakers measure well along any axis except directly on-axis. The waveguide has a very narrow hole right on axis because it is circular and the wavefront is so coherent (any other shape but circular or a less than ideal wavefront and this hole gets blurred.) If on-axis is desired then a crossover change can "modify" the hole, but then the power response at that frequency has a bump. The hole is fixable, but since I do not recommend on-axis listening it ends up being a non-issue. (In a three channel system the center channel has to be on-axis so the crossover gets modified.)
Earl,
Now you have my curiosity up!
This hole you speak of could you define a little more technically than just the word hole? What is happening in the frequency response or phase on axis and how wide an angle are you talking about? Also would you have this exact same phenomena in one of your elliptical waveguides on axis or is this only a problem with the symmetrically round waveguide? 
ps. I am assuming that these are exponential or thereabouts that expansion rate.
Now you have my curiosity up!
This hole you speak of could you define a little more technically than just the word hole? What is happening in the frequency response or phase on axis and how wide an angle are you talking about? Also would you have this exact same phenomena in one of your elliptical waveguides on axis or is this only a problem with the symmetrically round waveguide? ps. I am assuming that these are exponential or thereabouts that expansion rate.
Markus
Its fixable without acoustic tradeoffs, but the waveguide has to get larger and thats a big tradeoff from a marketing standpoint.
Kindhornman
Apparantly you are not up to speed on my waveguide work. My waveguides are not "exponential or thereabouts" they are Oblate Spheriodal. You will not see this effect on any waveguide except OS.
The "hole" is a few dB dip directly on axis and typically less than about 10° wide. Its "Q" and frequency location depends on the size of the waveguide and the radius of the mouth flare. A circular OS waveguide has all of the diffraction from the mouth edge arrive on-axis with exactly the same phase and so, at some frequency, the cancellation is deep and sharp. For an elliptical mouth the phases are not all the same and what happens is the dip gets much shallower and wider and in fact can even begin to disappear.
Other devices do not exhibit this phenomina because the wavefronts are not coherent enough and the mouth diffraction does not ever have the precise phase relationship that causes the cancellation.
On my website in the Polar Map App, there is a piston in a sphere example. This is a worse case for whenh the mouth is completly coherent and there is no flare. The axial dips can be seen to be quite deep and narrow and it is also obvious why they happen - the diffraction arcs from the edge of the disk meet on axis at periodic intervals of cancellation and reinforcement, much like a standing wave. It can be shown mathematically that this results form the wave being diffracted at the edge of the disk and rapping arround the sphere and re-radiating as it wraps.
Its fixable without acoustic tradeoffs, but the waveguide has to get larger and thats a big tradeoff from a marketing standpoint.
Kindhornman
Apparantly you are not up to speed on my waveguide work. My waveguides are not "exponential or thereabouts" they are Oblate Spheriodal. You will not see this effect on any waveguide except OS.
The "hole" is a few dB dip directly on axis and typically less than about 10° wide. Its "Q" and frequency location depends on the size of the waveguide and the radius of the mouth flare. A circular OS waveguide has all of the diffraction from the mouth edge arrive on-axis with exactly the same phase and so, at some frequency, the cancellation is deep and sharp. For an elliptical mouth the phases are not all the same and what happens is the dip gets much shallower and wider and in fact can even begin to disappear.
Other devices do not exhibit this phenomina because the wavefronts are not coherent enough and the mouth diffraction does not ever have the precise phase relationship that causes the cancellation.
On my website in the Polar Map App, there is a piston in a sphere example. This is a worse case for whenh the mouth is completly coherent and there is no flare. The axial dips can be seen to be quite deep and narrow and it is also obvious why they happen - the diffraction arcs from the edge of the disk meet on axis at periodic intervals of cancellation and reinforcement, much like a standing wave. It can be shown mathematically that this results form the wave being diffracted at the edge of the disk and rapping arround the sphere and re-radiating as it wraps.
I am NOT using an elliptical cross section, I would have thought from the discussion that was obvious. "Expansion rate" is a holdover from Webster which is an incorrect formulation of the problem. Oblate Spheriodal is the shape of the bounding surface.
The equation for this surface has a radius y at any point x along the central axis and a given throat radius of y0 and a final angle of theta is
y(x) = sqrt( y0^2 + x^2 Tan(Theta)^2 )
What size are we talking about?
P.S. It's "phenomena" not "phenomina"
I'm going to stop talking to you if you don't stop picking on me
The Summa, which is about 20 inches wide has almost no hole on axis. With some tricks that I have learned since that design I could reduce it to virtually non-existant. But going that route boosts the cost dramatically.
I'm going to stop talking to you if you don't stop picking on me
Sorry, no offense meant.
So, 20" min?
Markus,P.S. It's "phenomena" not "phenomina"
you are wrong. It must be "this phenomenon". Phenomena would need a "these" as pronoun.
Me too because I think it sounds more natural. And I think that there is something that is too narrow for living rooms.I prefer controlled wide dispersion.
So the question is also, what is wide and what is narrow.
Certainly, directional bass is much better than monopole bass if we are talking about woofers that are integrated in the main speakers.
Directivity is good as long as it remains fairly constant bottom to top.
Markus has obviously decided to become something that we call "Züzi" in our area and in the rest of Switzerland. That is yet anyther type of person, which is neither Swiss nor German. But I also need to put a bigDitto to that!!
here !I'd say it depends on what type of reproduction you're after. If "realism" (some call it "spatial effects") is the goal then certain room reflections need to be present and they also need to have certain properties.
If the most accurate stereo reproduction is the goal then it doesn't matter much what the directivity is as long as the reflections are sufficiently delayed and dimmed. Of course certain dispersion characteristics are better suited to reach that goal than others.
So directivity to me is not a goal in itself, it's just a property a device has to have. There are other very important variables that haven't been discussed yet. In what kind of room does the speaker need to work? How is everything set up (distance to walls)? How many listeners?
Markus has obviously decided to become something that we call "Züzi" in our area and in the rest of Switzerland. That is yet anyther type of person, which is neither Swiss nor German. But I also need to put a big
And the worst part is that I was born and raised in Swabia
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.