A horn is an impedance converter. Basically converting pressure at the throat to velocity at the mouth. (velocity at the mouth due to the apparent low impedance of air, but is really pressure as well.)
Just lite a transformer.
A waveguide does the same but also takes into account the propagation if the pressure wave throughout the guide/horn and the desired shape of it when it leaves at the mouth.
One can balance the propagation properties vs efficiency and impedance matching.
That is at least how I see it. A horn is a waveguide if you start looking at propagation and care about shaping it.
Just lite a transformer.
A waveguide does the same but also takes into account the propagation if the pressure wave throughout the guide/horn and the desired shape of it when it leaves at the mouth.
One can balance the propagation properties vs efficiency and impedance matching.
That is at least how I see it. A horn is a waveguide if you start looking at propagation and care about shaping it.
I view the whole "Horn vs. Waveguide" thing as a continuum. It's a 3 dimensional problem with diffraction, size, and constant directivity. Geddes and most other "waveguides" are on the extreme side of low diffraction and controlled directivity. LeCleach, Tractrix, and some other variants are low diffraction but more optimized towards size (for a given bandwidth) at the cost of constant directivity. There are Size/Directivity optimized horns like we often see in P.A. apps- diffraction throats on biradial horns or the like.
Sonically, in my experience, diffraction is pure bad, and so I would optimize on a 2 dimensional line. I also like the idea of maintaining CD, so my opinion is heavily towards the geddes solution, but I'd consider the LeCleach---Geddes "line" all "good" horns if executed well.
This is all prefaced with the fact that I'm fairly raw so far as horns are concerned. I've read a lot of theory but have thus far only built 1 system with a horn anywhere but as a supertweeter.
Sonically, in my experience, diffraction is pure bad, and so I would optimize on a 2 dimensional line. I also like the idea of maintaining CD, so my opinion is heavily towards the geddes solution, but I'd consider the LeCleach---Geddes "line" all "good" horns if executed well.
This is all prefaced with the fact that I'm fairly raw so far as horns are concerned. I've read a lot of theory but have thus far only built 1 system with a horn anywhere but as a supertweeter.
hI,
The difference is pure semantics.
In oldspeak they are both FNORD "Horns".
The newspeak term "waveguide" is used for marketing to distinguish a Horn aimed at pattern control first from a FNORD horn aimed at gain first.
As I am FNORD oldworld and speak FNORD oldspeak I normally prefer to call a spade a spade and a FNORD horn a horn...
Ciao T
The difference is pure semantics.
In oldspeak they are both FNORD "Horns".
The newspeak term "waveguide" is used for marketing to distinguish a Horn aimed at pattern control first from a FNORD horn aimed at gain first.
As I am FNORD oldworld and speak FNORD oldspeak I normally prefer to call a spade a spade and a FNORD horn a horn...
Ciao T
Hi,
An oblique reference to the late Robert Anthony Wilson - Illuminati Trilogy especially...
More at the home of all Fnord's, wikipedia:
Fnord is the typographic representation of disinformation or irrelevant information intending to misdirect, with the implication of a worldwide conspiracy.
Ciao T
An oblique reference to the late Robert Anthony Wilson - Illuminati Trilogy especially...
More at the home of all Fnord's, wikipedia:
Fnord is the typographic representation of disinformation or irrelevant information intending to misdirect, with the implication of a worldwide conspiracy.
Ciao T
That was the first hit on google and made some sense... but I thought it a little lightweight for you, fella!
BTW: welcome back. Have you exited the industry? Care to point me towards your work?
Dear Steve,
That is BY FAR overstating the real case.
I work for AMR and hold some shares, but I am only one of several people "behind" the company, that is before counting our dedicated staff, be Engineering, QA/QC or Logistics and Public Relations. I do have a bit to do with the development of the products...
Ciao T
That is BY FAR overstating the real case.
I work for AMR and hold some shares, but I am only one of several people "behind" the company, that is before counting our dedicated staff, be Engineering, QA/QC or Logistics and Public Relations. I do have a bit to do with the development of the products...
Ciao T
That is BY FAR overstating the real case.
I work for AMR and hold some shares, but I am only one of several people "behind" the company, that is before counting our dedicated staff, be Engineering, QA/QC or Logistics and Public Relations. I do have a bit to do with the development of the products...
Ok, ok. So you're just one of the company's behinds.
se
This is entirely wrong in fact, but is true given what has become of the term.
Here are the facts:
I have studied "horns" for decades and in the 1980's sometime I came to realize that it was impossible to truely design a horn for a specific directivity because the "Horn Equation", from which all "horns" derive, was one dimensional and as such could never be used to describe the three dimensional nature of the sound radiation from the device. It was adequite for impedance calculations, but not for directivity.
In 1991 I published a paper entitled Acoustic Waveguides. I used the term "waveguide" quite deliberately to differentiate this new approach from the "horn" approach. The two approaches do not yield the same results and devioces derived from these approaches tend to look and perform differently. It can also be hard to correlate characteristics from one into the other - like "cutoff", or "flare rate".
So, strictkly speaking, a waveguide is a device that uses one of the eleven coordinate systems that were defined in the original paper. The most common being the Oblate Spheriodal or OS.
The waveguide equations are exact, while the horn equation is, at best, an approximation, some being better than others.
Hence all waveguides are horns, but not all horns are waveguides. A waveguide needs to be designed with directivity as a fundamental characteristic, something that the horn approach cannot do. Horns are simply loading devices where the directivity is not predictable and hence it cannot be said to be a design criteria.
The term "waveguides" has however taken on a life of its own and I have seen it applied to just about anything, making the term meaningless in the context of marketing.
But originally there was a very clear distinction between a horn and a waveguide and in my mind there still is.
Dear Mr. Geddes,
Having worked with quite large scale sound reinforcement systems in the 1980's I distinctly remember the "horns" of the day and also many of the earlier devices.
While not backed as solidly by math as today; the better devices exhibited clearly designs with controlled and even directivity as design goal and usually as substantially achievement (if not as accomplished as your "waveguide" they nevertheless where a serious advance in the the state of the art). In fact, without this "CD" approach the larger concert sound systems of the day would have been near unworkable.
Of course these devices usually where called CD-Horns in the trade (Constant Directivity Horns) ignoring the often funny/annoying trade-names manufacturers attached to them and distinguishing them from many earlier designs that had flat gain/frequency response on axis, but widely varying response off axis.
Now I guess I am being curlish here, denying that "Waveguides" are a seperate category, distinctly different, from horns. This especially given that in a different forum I just declared a class struggle to get the "Class A1.5 Amplifier" terminolgy accepted I feel Class A1 and Class A2 leave too much confusion...
So maybe we can agree that:
By the defintion of your 1991 Paper a Waveguide(G) is a Horn, which has been designed according to certain specific and strict mathematical strictures.
By the common definition a Waveguide(C) is a horn designed to promote constant directivity (e.g. a CD-Horn).
By the common definition a Horn is a Horn.
So one may suggest that "Waveguides" no matter if Waveguide(G) or Waveguide(C) form one subclass of Horns.
One may also suggest that many other such subclasses are in eXisTenCe, such as the old multicell types and spherical wavefront horns (which despite their late 1930's german origin look awfully like 1990's "Waveguides(G)") and so on.
Waveguides, even according to your comments above are horns. Just one special kind of horns.
So:
Waveguide = Horn
Check that.
Yet:
Horn <> Waveguide
Well, I think we scratch the second unequation, as it contradicts the first agreed upon one.
QED.
Do you (in principle) agree?
We may still have the Waveguide(G) as a special way of designing horns even so, just as we may argue that the Horn(LeCleach) is a way of designing horns that is special.
What I'd actually like to know, will you ever make "Waveguide(G)" versions of horns that have asymetrical coverage pattern similar to the old asymetric JBL Horns how found in the Everest and now again popular in some sound reinforcement speaker horns (Nexo comes to mind)?
I found these devices in the 1980's era JBL Speaker absolutely amazing regarding image stability when moving away from the "sweet spot", just like it would be the case with a real orchestra (the violins do not shift position if you do, in a real concert hall), though seeming to be not entierly uncoloured sounding (sadly). Perhaps, as a result of using a diffraction slot between horn and driver.
This image stability thing is something current axis symmetrical; oval and rectangular constant directivity horns do quite poorly, including the ones I use.
Ciao T
Having worked with quite large scale sound reinforcement systems in the 1980's I distinctly remember the "horns" of the day and also many of the earlier devices.
While not backed as solidly by math as today; the better devices exhibited clearly designs with controlled and even directivity as design goal and usually as substantially achievement (if not as accomplished as your "waveguide" they nevertheless where a serious advance in the the state of the art). In fact, without this "CD" approach the larger concert sound systems of the day would have been near unworkable.
Of course these devices usually where called CD-Horns in the trade (Constant Directivity Horns) ignoring the often funny/annoying trade-names manufacturers attached to them and distinguishing them from many earlier designs that had flat gain/frequency response on axis, but widely varying response off axis.
Now I guess I am being curlish here, denying that "Waveguides" are a seperate category, distinctly different, from horns. This especially given that in a different forum I just declared a class struggle to get the "Class A1.5 Amplifier" terminolgy accepted I feel Class A1 and Class A2 leave too much confusion...
So maybe we can agree that:
By the defintion of your 1991 Paper a Waveguide(G) is a Horn, which has been designed according to certain specific and strict mathematical strictures.
By the common definition a Waveguide(C) is a horn designed to promote constant directivity (e.g. a CD-Horn).
By the common definition a Horn is a Horn.
So one may suggest that "Waveguides" no matter if Waveguide(G) or Waveguide(C) form one subclass of Horns.
One may also suggest that many other such subclasses are in eXisTenCe, such as the old multicell types and spherical wavefront horns (which despite their late 1930's german origin look awfully like 1990's "Waveguides(G)") and so on.
Waveguides, even according to your comments above are horns. Just one special kind of horns.
So:
Waveguide = Horn
Check that.
Yet:
Horn <> Waveguide
Well, I think we scratch the second unequation, as it contradicts the first agreed upon one.
QED.
Do you (in principle) agree?
We may still have the Waveguide(G) as a special way of designing horns even so, just as we may argue that the Horn(LeCleach) is a way of designing horns that is special.
What I'd actually like to know, will you ever make "Waveguide(G)" versions of horns that have asymetrical coverage pattern similar to the old asymetric JBL Horns how found in the Everest and now again popular in some sound reinforcement speaker horns (Nexo comes to mind)?
I found these devices in the 1980's era JBL Speaker absolutely amazing regarding image stability when moving away from the "sweet spot", just like it would be the case with a real orchestra (the violins do not shift position if you do, in a real concert hall), though seeming to be not entierly uncoloured sounding (sadly). Perhaps, as a result of using a diffraction slot between horn and driver.
This image stability thing is something current axis symmetrical; oval and rectangular constant directivity horns do quite poorly, including the ones I use.
Ciao T
Current TOTL JBL - K2, Array, Everest II - are all biradials. Pretend they suck, if you like.
Properly deployed to crossfire in front of the listener, most constant directivity horns/waveguides will generate an enhanced image rendition zone. Purpose-designed asymmetric defined directivity devices such as those used in JBL Everest I, S2600 and S3100 did quite a good job of it back in the 80s, but center channel sells more speakers....
Current TOTL JBL - K2, Array, Everest II - are all biradials. Pretend they suck, if you like.
Properly deployed to crossfire in front of the listener, most constant directivity horns/waveguides will generate an enhanced image rendition zone. Purpose-designed asymmetric defined directivity devices such as those used in JBL Everest I, S2600 and S3100 did quite a good job of it back in the 80s, but center channel sells more speakers....
I'll be happy to pretend that they suck! Having not heard them in such a way as to be able to evaluate, I'll reserve judgement. I consider the diffraction component to be a bad thing, but am willing to consider that it may not be a dominant bad thing. I certainly enjoyed my time with 2370s which are certainly diffraction horns.
Presently I'm using the PT "waveguides" while working on some OS waveguides. I need to scribble down some part values and provide the AK group with a schematic! I've got reasonably flat response through the horn passband. (1500-10k)
"Ever" - that's a long time. I have plans to make one, even have the mold, but the more I look into this the less attractive it seems to be. A lot of extra expense for no obvious gain and as I look out to the future all I can see is ever increasing costs and prices. Adding to this problem does not seem to be a good idea.
The image that you describe is precisely the kind of thing that so many comment about with my designs - but the OS waveguide is "colorless" owing to the lack of any diffraction.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.