Member
Joined 2003
I can't see the figures from here, but there is no technical discussion of the concepts. Only "this is what we get without and this is what we get with". Not much to go on. It appears that they are trying to get as wide a coverage as possible, quite the opposite of what I try and do. So perhaps the domes act as diffractors. I'll look at the figures when I get back to the US.
On constant directivity, one of the two primary benefits as I see it is removing the off axis sound to reduce early reflections. Now, am I behind here or is that the whole purpose of the star shaped felt cutouts we sometimes see around tweeters, and could this be done for woofers as well to control the midrange?
Woofers generate wavelengths that are too long to be effectivily directed. Only marginal directivity control below 500 Hz is at all posible even with very large speakers - by home speaker standards. Above 1000 Hz is really the only range at which any kind of effective control can be achieved.
I think that most of these "add-on" features are from hueristic developments - that's certainly true of the Klipsch design since no theoretical justification is given at all in the patent. I suspect that in general that is true for most of these types of modifications. They may work in specific cases, but it would be hard to extrapolate the concepts broadly enough to yield true engineering guidlines.
I should mention that this is in stark contrast to "waveguides" which started out as a purely theoretical investigation and then evolved into practical applications.
I think that most of these "add-on" features are from hueristic developments - that's certainly true of the Klipsch design since no theoretical justification is given at all in the patent. I suspect that in general that is true for most of these types of modifications. They may work in specific cases, but it would be hard to extrapolate the concepts broadly enough to yield true engineering guidlines.
I should mention that this is in stark contrast to "waveguides" which started out as a purely theoretical investigation and then evolved into practical applications.
I have found many patents that are in this category, which raises the question: If the patent's original purpose cannot be acheived unless an unrevealed condition is met, is it still a valid patent?
What frequency and waveguide/horn size do you consider as a limit for home reproduction?
Does it matter that there is little directivity control at the lower frequencies? Bass is generally assumed to be omnidirectional and if low enough in frequency then also non stereo.
If we want a horn/waveguide to perform down to, say 100Hz, what sort of impractical constraints need to be achieved to get this?
My 15" waveguide which is about 18" in diameter just gets down to 800 Hz, but below that the directivity control is collapsing. This is a big waveguide IMO, but not as big as they get.
There is no such thing as directivity in the modal region of the room so it can't be important that low. I believe its importance falls below about 500 Hz because of the way our hearing processes sound it can no longer really sort out reflections etc. Basicall the room is modal to about 200 Hz - so forget about directivity - and between 200 Hz and 500 Hz our hearing is not very accute. Our hearing starts to become extremely prcise at about 1 kHz and peaks at about 3 kHz. It falls above that.
Directivity control to 100 Hz would require a mouth that is about 12 feet across.
You may want to look at the Pi 7 corner horn. I don't know all the details, but I think by using a 24 inch horn, and the walls of the room, it has controlled directivity down to maybe ~300 or 400 Hz.
Pi Speakers - unmatched quality and state-of-the-art performance
The listening room in my house is bigger than the apartment I lived in a few years ago, and it's *still* too small for these speakers. Believe me, there's a point of diminishing returns when it comes to waveguides. Nobody needs a loudspeaker that's so large that you have to buy a new house to listen to them.
Calling a speaker in a corner a 90 degree waveguide is somewhat shaky IMO. I only consider the concept of "directivity" as being free field. To me, it is not applicable to the situation that you describe, because then anything placed in a corner is 90 degree CD. But there is a price to pay for this and they are called HOM. The HOM content from a corner loaded situation would be extremely high, making this not only less than desirable, but, again IMO, something to be completely avoided.
Earl, can the HOMs be seen using a velocity microphone (or microflown) inserted into a waveguide to sense side-to-side wave energy? I'm wondering what this would look like, with something like MLS or chirp drive. Or would putting the microphone into that area cause too much disruption itself?
Hi Bill
Not only would it work, but its been done. Makarski at Achen, did this with a horn. He measured very distinct not uniform motion in the wavefront. However, because this motion did not have a strong influence on the polar patern he concluded that it was not important. But wait! Says I. Not having a significant influence on the polar response and not be audible are distinctly different things. I think this is why the paper was only ever a preprint and not a full paper. He knew that he could not draw any audibility conclusions from what he found, but he clearly showed HOMs in action. It was only a few years later that Lidia and I published (although we already had the data) how small time perturbations - like HOM - could be audible well below the point of actually having a notable effect on the polar pattern. And most significantly - that these effects become MORE audible at higher SPLs. Can you say "horn harshness"? - very subtle time delayed echos, that can be quite annoying.
To me, the case is pretty much closed. Although the ney-sayers still persist to this day with their wild unsubstantiated beliefs.
Not only would it work, but its been done. Makarski at Achen, did this with a horn. He measured very distinct not uniform motion in the wavefront. However, because this motion did not have a strong influence on the polar patern he concluded that it was not important. But wait! Says I. Not having a significant influence on the polar response and not be audible are distinctly different things. I think this is why the paper was only ever a preprint and not a full paper. He knew that he could not draw any audibility conclusions from what he found, but he clearly showed HOMs in action. It was only a few years later that Lidia and I published (although we already had the data) how small time perturbations - like HOM - could be audible well below the point of actually having a notable effect on the polar pattern. And most significantly - that these effects become MORE audible at higher SPLs. Can you say "horn harshness"? - very subtle time delayed echos, that can be quite annoying.
To me, the case is pretty much closed. Although the ney-sayers still persist to this day with their wild unsubstantiated beliefs.
Ok, Earl, thanks. I might give it a try. I ran across a biological paper where some people researching insect sounds made a velocity microphone by essentially cutting the back off a cheap electret condenser mic capsule, and had good results. That doesn't seem like it should be overly difficult (though it might take destroying a small handful of capsules before one survives the trauma).
The better ones use a hot wire and measure resistance changes. Almost no effect on the velocity field. Your approach might be very insensitive for a sensor small enough to not destroy the field you are trying to measure. Depends on the frequency range you want of course. Makarski went up to 10 k I believe. That takes a pretty small sensor.
I some recall that in one of your papers you mentioned that absorptive surface wave guide might help reduce reduce HOMs. If we had sufficient frequency dependant absorption, wouldn't it be beneficial for usng such corner?
Lets say that the walls were somehow made entirely absorptive. Then the situation is indistinguishable from a free field, and there is no "waveguide" effect. So the whole discussion is moot. Reality will lie somewhere between reflective walls, a waveguide effect and HOMs, or absorptive walls, no waveguide effect and no HOMs. What is not possible is a waveguide effect and no HOM.
I disagree. There will still be directionality applied relative to free field. It won't be the same as a waveguide, most specifically it won't have the axial response track power response, but "Waveguide effect" must incorporate directionality in its definition or it's meaningless. After all, the elimination of boundary effects is a big part of why people use horns and waveguides in the first place.