|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
63.87.108.184
In Reply to: Re: overlooked topic posted by lecleach@musee.ensmp.fr on January 31, 2007 at 04:32:20:
Bonjour Jean-Michel,Thanks for the additional information and links.
The high frequency on-axis hole in John Sheerin's simulation is something that I have seen before, but depicted in a different format.
Earl Geddes told me that a round horn or waveguide will always have an on-axis hole in the frequency response. He said it has to do with an axi-symmetric reflection, and is not there in rectangular or elliptical devices.
You can see this sort of on-axis hole by clicking on the link below and scrolling down to page 11. This is a speaker that uses an oblate spheroid waveguide, which is an axi-symmetric device. There you will see a family of off-axis curves, which gives a somewhat less complete, but easier-for-the-layman-to-understand, presentation of the information contained in a polar map.
The curves are taken at 7.5 degree intervals. Notice that the on-axis curve dips below the first few off-axis curves in the range between about 4.5kHz and 9 kHz. I think this is the same phenomenon shown in John Sheerin's simulation, which doesn't go high enough in frequency to show that the on-axis hole closes back up at higher frequencies.
I also have measurements taken of a much smaller round waveguide than the one in Earl's paper, and it shows a similar on-axis hole in the response but at a higher frequency.
Anyway, I prefer to listen to round horns and waveguides from slightly off-axis, as the response is usually smoother there than directly on-axis. Again John Sheerin's simulation doesn't go up high enough to fully justify that conclusion regarding the LeCleac'h horn, but it looks like the trend is there.
Best regards,
Duke
Follow Ups:
Hello Duke,The strange thing about what you said about Earl Geddes's Summa is that Earl Geddes developped his waveguide in order to avoid those HOMs that seemingly are at the origin of the hole you mentionned.
For what it seems the waveguide shape succeed only partly to reduce those HOM. I think this is one of the reasons why, more recently, Earl Geddes filled entirely the waveguide with some foam.
Did the mesurement is done with the foam or not? If the answer is yes, most probably the origin of the hole on axis in the directivity response is yet to find...
Best regards from Paris,
Bonjour Jean-Michel,It is my understanding that the on-axis hole is an artifact of a virtually inevitable mouth reflection. It would take an infinite radius "lip" at the mouth to totally eliminate it. A larger radius "lip" does produce a measurably smaller on-axis hole.
The measurements I linked to are with the HOM[higher order mode]-absorbing refractive foam insert in place. The on-axis dip isn't a HOM artifact, so design features that address HOMs have no effect on it.
The Summa is designed to be listened to somewhat off-axis, which has other benefits aside from smoother first-arrival sound - but that's another story.
A la prochaine,
Hello Duke,IMHO: in the case of a waveguide filled by foam, the origin of such a hole of 3 DB cannot be a reflection of waves from the mouth.
Best regards from Paris,
Jean-Michel Le Cléac'h
In my opinion you are incorrect.No mouth can be made such that there is no diffraction and as such this diffraction will meet at the axis in anti-phase at some point. This causes a hole. It is not an HOM effect and is unaffected by the foam. A larger mouth radius makes it smaler and a larger mouth makes it go lower in frequency.
I have tested this all and it is quite correct.
I have also stated numerous times that no waveguide has NO HOM, only some have a lot less. The drivers create HOM even if the waveguide does not.
Duke is dead right in what he says. He has learned his lessons well.
Earl Geddes
Hello Jean-Michel,What do you believe causes the on-axis response hole?
Do you believe it has one cause in the case of your horn (as modelled by John Sheerin), and another cause in the case of Earl's waveguide?
Merci,
Duke,As I told you I don't see myself such a hole on my measurements some of them can be seen at:
http://www.musique-concrete.com/mesures/321direct.gif
due to the precision of the measurement, if this axial hole
exists it is far less than 0.5dB with my horn.I don't want to criticize John Sheerin simulations, there are very enlightning about differences between horn shapes. A limitation to those simulations is that they are only performed in 2D not 3 D (someone will surely say this has a minor influence in the case of an axial horn...)
You know that the profile of horns calculated with my method is curving back at the mouth. Diffraction is therefore reduced. This explains both the very pure wave shape on impulse reponse and the very smooth electric impedance curve.
Most waveguides are cut too sharply at the mouth and for sure diffraction will occur with more evidence.
Best regards from Paris
Jean-Michel Le Cléac'h
This post is made possible by the generous support of people like you and our sponsors: