gedlee said:
I used about four drops of glue per layer, and it's at the edges. While that's not ideal, the drops of glue are small enough that it should be relatively transparent. Putting the glue at the edges should minimize the effect also.
Because any obstruction at the throat is a problem, there is no glue at all in the last inch or so. It's held into the waveguide by compression.
If I were to do it again, I would use thread, and just sew the whole thing together.
Another option that worked decently is to wrap the whole thing in nylon; a pair of pantyhose for a buck worked quite well.
Earl what would be a good way to measure HOM's? It strikes me that a gated MLS signal may not be the best. Perhaps an analog swept sine? Anything to look for in the signal?
I'll let this get back on topic, but to those who complimented my photos, thank you, much appreciated. I actually have always thought I was a very amateur (at best) photographer. I took classes at an art school as a kid, kept up with it, have belonged to clubs and art organizations, but mostly because of my Mom's own art work, and her involvement in the arts. My involvement was always largely to support the arts, not to be a part of it. It wasn't until I was approached by some of the judges for our local Art Museum's art festival, where I as asked to submit some work, that I began thinking their might be interest. This was followed by a few nice bids on some pieces, and that was that. Unfortunately, graduate school has taken it's toll on my free time, and while I have stuff on display out here in Indiana, as well as in New York, I've just not been actively shooting anymore, nor have I worked to sell anything.
augerpro said:
There is, I believe, some discussion of the presence of HOM in the impulse response in a power point presentation on my web site. I don;t think that they would be visible from a frequency response plot, but do appear to be visible in the impulse response.
Hello Mr. Geddes,
i am interested in how you deal with the narrowing of directivity in the OS waveguide at the lower end of its frequency bandwidth, which also happens with the conical flare. Tom uses a higher flare rate for what seems to be about the last 1/5 of the length of his synergy horns to counter this (on a different board he wrote it should be 1/3, but it looks definitely shorter). Does the mouth radius in your waveguides have a similar effect or are you somehow compensating for it in the crossover?
Thank you 🙂
i am interested in how you deal with the narrowing of directivity in the OS waveguide at the lower end of its frequency bandwidth, which also happens with the conical flare. Tom uses a higher flare rate for what seems to be about the last 1/5 of the length of his synergy horns to counter this (on a different board he wrote it should be 1/3, but it looks definitely shorter). Does the mouth radius in your waveguides have a similar effect or are you somehow compensating for it in the crossover?
Thank you 🙂
Member
Joined 2003
Hi Earl,
Would you be more specific about your measurement technique? I, for one, am very interested in the subject and would like to have a reliable way of measuring progress. I have seen "funnies" in the impulse response, but nothing repeatable or clearly attributable to HOMs. In this post you mention you have measured them, and at a high level, so I think they would be quantifiable with the right measurement.
MaVo said:
The mouth radius is the main thing that helps, but the reality is that nothing is going to elliminate it. What you need to do is operate the waveguide above this frequency.
Paul W said:
I just looked at the impulse response with and without foam and the later parts of the impulse response were attenuated by the foam. This is not hard proof that they were HOM, but it is consistant with the concepts.
The HOM are large compared to cabinet radiation, but not necessarily compared to total output, except at perhaps certain frequencies. Diffraction can be a very difficult thing to measure since you cannot ever measure the diffraction part itself only its effect on the main radiation.
Hi Earl
to no longer hitch hike Lynns thread – may I ask you some more questions about HOM ?
– seems it is the only way to find out what's all the fuss about HOM as you seem not to be willing to give a direct explanation nor measurement advice for one reason or another - and also seem to be not happy with Bjorns definition on the HOM subject .
😉
OK – ready for a little bit more of question and answer gaming?
🙂
1. is there something like "lower order modes" in contradiction to HOM (higher order modes) - or - what exactly is *modes* in this context?
2. can I look at HOM as a "defect" in the sound field that's not stable over time at a certain place in space?
Michael
to no longer hitch hike Lynns thread – may I ask you some more questions about HOM ?
– seems it is the only way to find out what's all the fuss about HOM as you seem not to be willing to give a direct explanation nor measurement advice for one reason or another - and also seem to be not happy with Bjorns definition on the HOM subject .
😉
OK – ready for a little bit more of question and answer gaming?
🙂
1. is there something like "lower order modes" in contradiction to HOM (higher order modes) - or - what exactly is *modes* in this context?
2. can I look at HOM as a "defect" in the sound field that's not stable over time at a certain place in space?
Michael
I answered all your questions about HOM.
Bjorn never discussed HOM to my knowledge so how could we disagree?
The lowest order mode is the normal wave which propagate along the axis and is everywhere perpendicular to the walls. All other modes are higher than this, do not propagate along the axis and reflect off of the walls as they travel. The term "higher" refers to the fact that these modes have a "cut-in" effect which means that they canot exist below some frequency - the lowest order has a "cut-in" of 0 Hertz. The "cut-in" frequency gets higher and higher wil mode number. Mathematically these modes are eigenvalues and eigen functions of the Dif-EQ just as modes in a room are.
HOM are completly stable in time just as the main mode is. They are just other ways in which the wave can travel. The situation is completely analogous to HOM in tubes - look that up.
Bjorn never discussed HOM to my knowledge so how could we disagree?
The lowest order mode is the normal wave which propagate along the axis and is everywhere perpendicular to the walls. All other modes are higher than this, do not propagate along the axis and reflect off of the walls as they travel. The term "higher" refers to the fact that these modes have a "cut-in" effect which means that they canot exist below some frequency - the lowest order has a "cut-in" of 0 Hertz. The "cut-in" frequency gets higher and higher wil mode number. Mathematically these modes are eigenvalues and eigen functions of the Dif-EQ just as modes in a room are.
HOM are completly stable in time just as the main mode is. They are just other ways in which the wave can travel. The situation is completely analogous to HOM in tubes - look that up.
Regarding HOM measurements, it seems to be very important to verify that the impulse measurement is free of any glitches/artifacts. To my knowledge the most robust technique seems to be Angelo Farina's LogSpeep technique (often labelled ESS, Exponential Sine Sweep). The nice thing is that it also seperates the impulse in its individual distortion components (seperate IRs for fundamental and each harmonic).
Info here:
http://pcfarina.eng.unipr.it/Public/Papers/134-AES00.PDF
(for the impatient, see figs. 18 and 19)
http://www.anselmgoertz.de/Page10383/Monkey_Forest_dt/Manual_dt/Aes-swp.pdf
Alas I don't know which audio programs use this method, except for Farina's original Aurora plugin for CoolEdit.
- Klaus
Info here:
http://pcfarina.eng.unipr.it/Public/Papers/134-AES00.PDF
(for the impatient, see figs. 18 and 19)
http://www.anselmgoertz.de/Page10383/Monkey_Forest_dt/Manual_dt/Aes-swp.pdf
Alas I don't know which audio programs use this method, except for Farina's original Aurora plugin for CoolEdit.
- Klaus
Dr. Geddes is obviously the expert here on HOM's and can speak most intelligently about them, but this is my take on the measurement issue.
First, I think it needs to be cleared up that in physics, many things are found to likely exist due to the math. In the case of sound waves, math models sound quite well. This mathematical model is what, as I understand things, first drew attention to the HOM issue. This is not the same thing as theoretical, and I don't see there being an arguement over their existence in the scientific community, which I do see happening on this forum.
HOM's have been measured in certain wave systems, but as I understand it, never really directly measured in a precise fashion with speakers. When you think about what they are, and what effect they should have on the response, it makes sense that they couldn't be easily directly measured. HOM's are delayed energy, so what should they look like in a measurement, well they have to be derived, as mentioned, from the impulse response, since their effect is going to be delayed, and found in the phase. Because they are higher order, they will only exist in a higher frequency range, Dr. Geddes refers to this as the "cut-in". There for, when looking for them in measurements, you need to look at higher frequency delayed energy. You have to separate this from other potential sources of higher frequency delayed energy, and they need to be at a level that is greater than the noise in the impulse response. Given that the impulse response is a mathematically derived function, this ability can exist, but is not a common intent with measurement software. Most impulse responses are very noisy.
It's been my take from Dr. Geddes that he mostly uses methods to indirectly measure HOM's. The primary way he evaluated the issue was in the mathematics of his waveguide. The Foam may have been mathematically analyzed as well, but really, it's such an ingenious and simple idea, I could see not bothering (I would think adding the function of the foam into the equation somewhat complicated). HOM's travel down the waveguide or through the tube by bouncing around it in a specific way, therefor, they must travel a longer distance, hence the delayed energy. If you place foam into a tube, while everything travels a longer distance now, the HOM's would be exponentially longer. Given the open cell nature of the foam used and it's shape within the waveguide, it further makes sense just how much longer this path would be, since the longest straight path through the foam isn't the straight path out.
I think for the average person with normal measuring abilities, even using Ferina's method, the best we could expect to do is look for ripples in the impulse of the high frequency driver, and see if something like foam can reduce those. If they can, then then it fit's the model for HOM's and seems a reasonable explanation.
I would imagine that we could more precisely exam the HOM issue by knowing what frequencies they should occurr at within a given waveguide, tube, or horn, then gating the impulse respectively. Another possibility would be to look for ridges in the waterfall plot, but given how low in level they are, I would imagine that would be very hard to do.
First, I think it needs to be cleared up that in physics, many things are found to likely exist due to the math. In the case of sound waves, math models sound quite well. This mathematical model is what, as I understand things, first drew attention to the HOM issue. This is not the same thing as theoretical, and I don't see there being an arguement over their existence in the scientific community, which I do see happening on this forum.
HOM's have been measured in certain wave systems, but as I understand it, never really directly measured in a precise fashion with speakers. When you think about what they are, and what effect they should have on the response, it makes sense that they couldn't be easily directly measured. HOM's are delayed energy, so what should they look like in a measurement, well they have to be derived, as mentioned, from the impulse response, since their effect is going to be delayed, and found in the phase. Because they are higher order, they will only exist in a higher frequency range, Dr. Geddes refers to this as the "cut-in". There for, when looking for them in measurements, you need to look at higher frequency delayed energy. You have to separate this from other potential sources of higher frequency delayed energy, and they need to be at a level that is greater than the noise in the impulse response. Given that the impulse response is a mathematically derived function, this ability can exist, but is not a common intent with measurement software. Most impulse responses are very noisy.
It's been my take from Dr. Geddes that he mostly uses methods to indirectly measure HOM's. The primary way he evaluated the issue was in the mathematics of his waveguide. The Foam may have been mathematically analyzed as well, but really, it's such an ingenious and simple idea, I could see not bothering (I would think adding the function of the foam into the equation somewhat complicated). HOM's travel down the waveguide or through the tube by bouncing around it in a specific way, therefor, they must travel a longer distance, hence the delayed energy. If you place foam into a tube, while everything travels a longer distance now, the HOM's would be exponentially longer. Given the open cell nature of the foam used and it's shape within the waveguide, it further makes sense just how much longer this path would be, since the longest straight path through the foam isn't the straight path out.
I think for the average person with normal measuring abilities, even using Ferina's method, the best we could expect to do is look for ripples in the impulse of the high frequency driver, and see if something like foam can reduce those. If they can, then then it fit's the model for HOM's and seems a reasonable explanation.
I would imagine that we could more precisely exam the HOM issue by knowing what frequencies they should occurr at within a given waveguide, tube, or horn, then gating the impulse respectively. Another possibility would be to look for ridges in the waterfall plot, but given how low in level they are, I would imagine that would be very hard to do.
gedlee said:
Yes – thanks for that - hopefully you stay with this habit - but I didn't always understand what you wanted to tell me – or maybe you didn't always understood what I wanted to ask
😉
gedlee said:
No he didn't do a discussion – but he did a summarisation in his papers I was referring to in one of my questions – anyway – don't care – its not important anymore.
gedlee said:
Thanks for pointing to this – my search brought up articles about HOM suppression (at specific mode orders) in radar apps about HOM (generated by dipole and quadruple sources) in bore holes and for measuring HOM in bubbled water and a lot more – it will take some time to filter out what is of interest here.
As I understand for now there is no way that a wave front propagates other than on axis – for frequencies below so called cut off (*not* the one of the horn) - which depends on free space dimensions available perpendicular to axis with respect to the wavelength – above cut off non axial propagation is possible as well – with the impacts that time of flight is different and kind of mirror sources may be created in case of straight boundaries like my cylinder horn
1. what does HOM sound like – any specific pattern to distinguish from other abnormalities?
2. if HOM is to make the sound field (predictable or at least stable) defective like Swiss cheese - this is nothing different of any diffraction / scattering effects – mere comb filtering due to overlay. So then, are there other effects included that make HOM "number one enemy" for you?
Michael
gedlee said:
Speaking of this, one eerie thing that I've noticed with the Summas is they're VERY sensitive to obstructions in front of them. I'm guessing this is because they're CD, so the treble energy is "fanned out" across a wide area?
The sensation is almost like what happens when someone walks in front of my projector. Except with sound instead of light.
It's a bit hard to describe, and I've never heard another speaker that does this.
But it means that anything between the speaker and the listener is a real problem. For a while I had the speakers near a 4" post in my listening room, and I had to move them away from the post, because it seemed to be screwing up the soundstage.
And it was a good 40 degrees off axis, on the other side of the speaker.
Thanks for your contributions, too, Michael.
That Dipole Horn thing, for example. Of course now I have no way to prove that the same basic idea was boggling my mind once I learned the Mundorf AMTs are true open back, more or less exactly one year ago (when I spotted a rear photo of the 2740).
Nevermind, though...
BTW, maybe much of the hype about of the original Heil AMT is due the fact that it also has some sort of waveguided dipole properties?
- Klaus
That Dipole Horn thing, for example. Of course now I have no way to prove that the same basic idea was boggling my mind once I learned the Mundorf AMTs are true open back, more or less exactly one year ago (when I spotted a rear photo of the 2740).
Nevermind, though...
BTW, maybe much of the hype about of the original Heil AMT is due the fact that it also has some sort of waveguided dipole properties?
- Klaus
mige0 said:
I think that HOM are what is generated in this test. http://www.audioheritage.org/vbulletin/showthread.php?t=12967
Easy to do, and hear with and without.
Diffraction is the root cause of HOM's. But since the HOM are diffraction that has been directed back into the horn, they can reverberate in a way that simple diffraction off of, say, an enclosure corner, does not. The diffraction into the horn, will come out, and repeat the process with diminishing energy. Think echo. Whereas diffraction that doesn't have this path will have a single cycle.* I think that this is what our ear picks up.
*Yes, I know that single vs. multiple cycles is somewhat of an forced qualitative distinction, as sound is reflected multiple times off of surfaces in the area. However, I would think that the horn will more efficiently direct this recycled energy toward the listener for the same reasons it's more efficient than a flat surface is for the original signal. I think the issue is that the horn makes these particular echos audible to our perceptual system.
As pjpoes implied, the current measurement tools may be too crude to allow us to reliably pick out the HOM's from the imperfect data - GIGO. However, we (that is, you guys with the smarts) should be able to model them. This way we can start out with a mathematically perfect input, i.e., no noise. We could at least then see what what they look like in various presentations or derivations, and their relative magnitude. Based on what I understand from these discussions, we'd expect them to on the order of the current noise contribution, or less.
Sheldon