Elias said:
Thanks Elias for linking the Michael Makarski papers – very revealing.
And saves me a lot of time as a lot of measurements and simus were done already – way better than I ever could do.
Although one thing isn't clear to me (well 99,99% of the math too 😉 ) – you are right – or at least I read it the same way - that Michael Makarski tells us that the horn itself does not contribute a lot to excite HOM - not even for the tough (no transition section) attached conical horn of obviously less than 45deg ½ angle.
Either he is way more clever (most probably) than me or his BEM simulation simply didn't account correctly for reflection due to diffraction at the driver / horn joint - or the compression drivers are really really bad HOM generators.
What I find most interesting is that the total energy he calculates for radiation in higher order modes is pretty low – on the other hand - we would see really weird (Swiss cheese) sound fields if this were not the case.
With the above in mind it is really interesting to compare the BEM simulated soundfields of Bjorn Kolbrek at JMLC thread - which compare the LeCleach with the OS.
All in all - lots of additional details – but not too contradictory to my low level sims and noob conclusions – this confirmation (as far as I can see it) makes me pretty happy!
Michael
Before one criticises the concept of HOM, one should read http://gedlee.com/Audio_trans.htm and other available papers and then discuss the matter on this base. Everything else probably disqualifies itself as http://en.wikipedia.org/wiki/Troll_(Internet) behaviour.
MaVo said:
To belive in Earls mostly semantic "HOM" concept is fine - but - please - don't call people with different experience or point of view "troll" - there is way more on the matter than what Earl likes to take into account
...
Also it is Earls "game" to play on the scientific level – hence – if the prove (if we can call it at all) isn't exactly convincing – there is criticism / correction *needed*.
That's simply the rules of the "science game".
Also a new term like "HOM" has to be linked to a specific result not explainable with the mechanisms already known (which isn't the case here) *if* we play the "science game" correctly – otherwise we only get Babylon .
Michael
Hello,
To summarise HOM, there is two condition required for HOM to exist (according to Makarski paper).
1) The driver must excite the higher order mode by loosing the pistonic movement of the membrane (i.e. break up resonance)
2) The excited mode has to be coupled to the horn.
After that the horn is just there as a transfer medium.
- Elias
To summarise HOM, there is two condition required for HOM to exist (according to Makarski paper).
1) The driver must excite the higher order mode by loosing the pistonic movement of the membrane (i.e. break up resonance)
2) The excited mode has to be coupled to the horn.
After that the horn is just there as a transfer medium.
- Elias
Hello,
Next thing we should do is to find out the ranging of importance of HOM generation mechanisms - driver break up vs. something else(?).
And we will find it! 😀
- Elias
Next thing we should do is to find out the ranging of importance of HOM generation mechanisms - driver break up vs. something else(?).
And we will find it! 😀
- Elias
Elias said:
This is the one part of the Makarski paper with which I take exception. I was a reviewer of the paper for the AES and I raised this issue and asked for proof of it (this was a peer reviewed journal where statements DO have to be proven or referenced). When one was not forthcoming the paper was rejected for publication which is why it is only available as a preprint (preprints are NOT reviewed).
There are three sources of HOM
1) diaphragm non-pistonic motion
2) phase plug and horn driver interface (which to me are the same thing)
3) the horn itself
In any given device with any given driver these three sources could come in any mixture and further this mixture would be frequency dependent. The fact that there is so much confusion about HOM is evidence of the difficulty in understanding them. I try to clear up these misunderstandings as they come along, but I really get tired of being harassed for this.
mige0 said:
Michael
I DO play the game correctly and the use of the term Higher Order Modes here is not only absolutely correct but completly consistant with all previous uses of the term, which are quite numerous and rigorous. It is your and others incorrect claims that HOM are nothing new and just diffraction and reflection that is incorrect.
It is you who are abusing science by making broad claims based on very flimsy data and expecting the rest of us to accept them as fact. That IS NOT the scientific method.
Hello Earl
Can I take this:
"1) diaphragm non-pistonic motion"
As a yes to this question in the HOM measurement thread?? Or does it apply to compression drivers only??
My question:
"So do dynamic drivers mimic what HOM's sound like when they are pushed hard and the break-up modes become audible?? Is that what we should be listening for?? Looking at the work that shows them at the mouth of the horn it reminds me of some of the break-up mode pictures I have seen made using a laser of dynamic speaker break-up modes."
Rob 🙂
Can I take this:
"1) diaphragm non-pistonic motion"
As a yes to this question in the HOM measurement thread?? Or does it apply to compression drivers only??
My question:
"So do dynamic drivers mimic what HOM's sound like when they are pushed hard and the break-up modes become audible?? Is that what we should be listening for?? Looking at the work that shows them at the mouth of the horn it reminds me of some of the break-up mode pictures I have seen made using a laser of dynamic speaker break-up modes."
Rob 🙂
Robh3606 said:
No, that is not correct. HOM for a mechanical structure are quite different than those for an acoustic system. The 3D nature of the acoustics problem allows for dispersion of the wave as it propagates in space. While the modes in a membrane are themselves also dispersive they are only so in frequency not in space (the wave speed is frequency dependent in a membrane, but it is spatially dependent in a waveguide - very different things).
The acoustic field in the mouth of the horn would indeed be quite analogous to the breakup modes of a membrane, that part is absolutely correct. The total motion being the sum of the primary mode and the HOM in both cases. But the difference comes from how that pattern comes about. The HOM in the diaphragm are not related to the HOM in the waveguide, but the opposite is true. For example, there will still be HOM and a complex acoustic pattern at the mouth even if the diaphragm is perfectly rigid with no HOM.
In a real system the audible issues will be a very complex mixture of the HOM in the diaphragm, those in the phase plug and those in the waveguide and each contribution depends on the others. This is exactly why it is so hard, quite possible impossible, to say that a certain sound quality does or does not stem from HOM in the waveguide.
BUT, here is some data that makes the issues at least somewhat clearer. 1) identical drivers on a poor horn will sound bad, hence the HOM in the driver are not dominate because 2) the same driver on a good waveguide sounds better, hence the HOM created by the waveguide or horn must be a strong contributing factor. Finally 3) adding foam does not change the creation of HOM by either the driver or the horn/phaseplug, but does improve the sound, hence the foam must have an effect on something that causes the poor sound quality. Is HOM reduction the whole story? I doubt it. Are the HOM contributed by the waveguide a factor, of that I have little doubt and no one that I know of has studied this as much as I have.
Some time ago I had tried some stuffing in font of the horn, and it does improve sound. I would expect foam to be a more elegant way of doing it. Whether there is a better way, I think is something we would always try to look for.
gedlee said:
I think that is the important part, at least for an end-user.
I think this is why people should start looking at CSD plots at varying spl's. Perhaps even more so for typical "cone" drivers.
Past a "cone" driver's pistonic operation and into deformation, membrane modes can actually shift with varying spl - especially lower mass diaphragms typical in "full-range" drivers. Also, this sort of behavior extends to typical mid-range cones (for average applications).
Subjectively however I've found that increased linear decay (unless specifically producing high amplitude narrow "q" resonances), does not introduce a "fatiguing" sound. The sound is basically less "clean" sounding, or less "transparent".
With shifting modes the sound becomes so much less "clean" sounding that imaging becomes "confused". Image localization becomes more difficult to pin-point, its physical character FAR less resolved (on really good systems), and all imaging becomes more "homogenized". People often attribute IM distortion to this, but I don't think IM distortion sounds like this, instead IMO it adds an unpleasant "greyish" sound that is fatiguing but does NOT change the "architectural" aspects of reproduction.
Scott
I do tend to agree with you about breakup modes in direct radiators, but I'll leave the subjective decriptions to others like yourself who feel inclined to such things. Myself, I, would never use a driver into the region of cone breakup. The first mode (rim resonance) is usually fairly easy to control with the proper surround and can actually be used to effect in the crossover itself, but beyond that no driver should ever be used - I don't. While this does show up in the CSD, but it is also quite evident in the classical polar response etc. meaning that CSD is not really showing anything that's not also evident in more typical data.
And this really has nothing to do with waveguides since, as I said, waveguide HOM and cone HOM have little to do with each other.
I do tend to agree with you about breakup modes in direct radiators, but I'll leave the subjective decriptions to others like yourself who feel inclined to such things. Myself, I, would never use a driver into the region of cone breakup. The first mode (rim resonance) is usually fairly easy to control with the proper surround and can actually be used to effect in the crossover itself, but beyond that no driver should ever be used - I don't. While this does show up in the CSD, but it is also quite evident in the classical polar response etc. meaning that CSD is not really showing anything that's not also evident in more typical data.
And this really has nothing to do with waveguides since, as I said, waveguide HOM and cone HOM have little to do with each other.
gedlee said:
Agree - as I only occasionally like to play the "scientific game" by myself...
🙂
gedlee said:
I agree that I didn't find it *strictly* incorrect what you have done by introducing "HOM" as a concept for audio apps as it might be "not completely useless" to have an incisive term available for that subset of mechanisms - just a little bit 😉 as you failed to link a prove for its specific relevance (and I still see non - even now) for horns.
But I'm definitely *not* with you that HOM isn't any more than a subset of "diffraction- reflection-delay-interference".
It does not matter if that happens in the driver – at the throat joint – along the horn contour - at the mouth - or from any other more or less axisymmetric diffraction points at any object.
If we again look at the images at
http://www.diyaudio.com/forums/showthread.php?postid=1791730#post1791730
we can see that the OS wave guide contour doesn't exactly provide a way smoother sound field.
I guess you agree that the level of raggedness (Swiss cheese) of the sound field is an ultimate measure of *any* "diffraction- reflection-delay-interference" involved.
gedlee said:
No – my simus are low level, sure – but way more revealing / clarifying on an intuitively level to most of us than a lot you (and others) have written about the subject over the years.
In addition to that – this few pix I've shown do actually provide *additional* aspects not included in any other simus, measurements or concepts of explanations I've seen so far.
Michael
Hello,
the paper is not a preprint
http://www.akustik.rwth-aachen.de/pub/mma/AES117paper.pdf
It was presented to the public at the AES 117th Convention
2004 October 28–31 San Francisco, CA, USA
- Elias
the paper is not a preprint
http://www.akustik.rwth-aachen.de/pub/mma/AES117paper.pdf
It was presented to the public at the AES 117th Convention
2004 October 28–31 San Francisco, CA, USA
- Elias
gedlee said:
Hello,
It's a pleasure to notice the number of sources of HOM has now increased from one to three 🙂
In order to continue within the subject of HOM, let's take a look at Makarski's AES paper (not preprint) to gain some insight how HOM is radiated out of the horn.
See figures 16, 17 and 18 from
http://www.akustik.rwth-aachen.de/pub/mma/AES117paper.pdf
Pay a particular attention of how narrow is the beamwidth of higher order modes. The are very narrow indeed. The are like needle pinching, or a laser pointing on axis. If you move 1 to 2 degrees off axis the HOM level is down 20dB.
That is, if you listen slightly off axis you will never hear the HOM 😀
- Elias
gedlee said:
It's a pleasure to notice the number of sources of HOM has now increased from one to three 🙂
In order to continue within the subject of HOM, let's take a look at Makarski's AES paper (not preprint) to gain some insight how HOM is radiated out of the horn.
See figures 16, 17 and 18 from
http://www.akustik.rwth-aachen.de/pub/mma/AES117paper.pdf
Pay a particular attention of how narrow is the beamwidth of higher order modes. The are very narrow indeed. The are like needle pinching, or a laser pointing on axis. If you move 1 to 2 degrees off axis the HOM level is down 20dB.
That is, if you listen slightly off axis you will never hear the HOM 😀
- Elias
Elias said:
"Papers" only appear in the Journal and are peer reviewed. "Preprints" are given at conventions and are not reviewed. The above was a "preprint".
Elias said:
I too noted this aspect of Makarski's work with great interest, but have not been able to pursue it any further.