I don't either, tho I've been told that it does. It seems to me that the hypotenuse or the diameter of the mouth is what matters.
The advantage of square or rectangular is a matter of ease of construction, isn't it? Not acoustic.
If that's the case, then I understand, but it is certainly not easier for me to do rectangular. It is the exact opposite. And it seems like the transition from round to rectangular is tough no matter how you do it.
Maybe flattened round shapes, like an oval, were difficult or expensive to manufacture at some point?
Might be interesting to convert a rectangular horn to a more rounded shape to hear and the measure the change.
Might be interesting to convert a rectangular horn to a more rounded shape to hear and the measure the change.
I have seen and measured the differences (but as you know I don't put too much stock in listening tests). Generally the directivity control fails sooner and more dramatically along the diagonals - basically the "best case" is nearly always the horizontal and vertical axis - the only ones that anyone ever shows. I recently measured a "popular"-"waveguide" and it was terrible along the diagonals. I simply could not understand what the designers were think when they made it square. In reality I think that it is just tradition and that there is no real reason to do it, just inertia (which is a powerful thing in audio.) As far as I can tell there no advantage to rectangular.
With a square one, probably not. With a non-square one, you could get a closer C-C for better vertical off-axis. Of course, the sacrifice is more pattern control loss on the vertical planes.
You miss the point - elliptical is the better choice than rectangular. I said that above. Circular is nothing more than a special case of elliptical. Elliptical is the best over-all choice. Rectangular offers nothing on elliptical. I have made elliptical waveguides.
I think people are used to thinking about area (and thus squares / rectangles) if they've been thinking about low frequency horns where loading is the primary goal instead of directivity.
How well does the "Peavey quadratic throat waveguide" work?
http://aaassets.peaveyelectronics.net/pdf/qwp1.pdf
http://aaassets.peaveyelectronics.net/pdf/qwp1.pdf
I covered the Peavey design years ago. It is virtually identical to an OS, they just wanted to claim that they had found something new, but basically it was just a copy.
Yes, I can see it is a copy of your design, and it has probably been discussed before in this thread.
I was more interessed in the transition from circular to square and how well it could work.
I was more interessed in the transition from circular to square and how well it could work.
There are three documents on the web describing the Quad Throat design (by Charles Hughes).
One is a white paper is essentially useless. The other two are similar (a patent and a AES conference presentation). These two provide more detail and some performance measures. Importantly, they outline a technique for the circular to rectangular transformation. There is also mention of the advantage of placing foam in the final third of the horn to help with the flare at the mouth of the horn.
I am not sure that a comment beyond the fact that square offers nothing is really appropriate. Its not worth going from round to rectangular, but if you have to do it then do it as slowly as possible.
Earl, one question about rectangular designs before ditching the idea out of this thread for good 
Is there any advantage in having a radial design for the vertical profile (like a TH4001 or arai A290, with a central symmetry originating "somewhere" around the phasing plug...) rather than a straight profile?
Is there any advantage in having a radial design for the vertical profile (like a TH4001 or arai A290, with a central symmetry originating "somewhere" around the phasing plug...) rather than a straight profile?
I don't follow the question.(Maybe because I don't know what a TH4001 or arai A290 are.)
Square waveguides are not in and of themselves a bad thing, but all good designs (such as a Prolate Spheriodal waveguide) would require a square aperature from the compression driver. The aperature from the driver dictates the waveguide design - that should be easy to understand.
Square waveguides are not in and of themselves a bad thing, but all good designs (such as a Prolate Spheriodal waveguide) would require a square aperature from the compression driver. The aperature from the driver dictates the waveguide design - that should be easy to understand.
Having used the Hughes type throat transition on a number of rectangular horns in small PA and stage monitor cabinets, found that the polar response is as good or better than similar horns using "pinched throats", and subjective sound quality is much better, probably from a reduction of H.O.M.s.
If keeping HF reflections off the floor and ceiling are important while maintaining a wide horizontal dispersion, a rectangular conical waveguide has definite advantages to a round conical, and is far easier to build than an ovaloid conical.
However, a correct smooth round to square or rectangular (or ovaloid) transition is a lot harder to build than a round to round waveguide.
Art
Member
Joined 2003
"ovaloid" - but I prefer "elliptical", is not all that difficult to make, but I have not as yet seen the benefits that make it worth doing. Lots of guesses as to why it should be beter, but when I analyze the situation I see lots of downsides as well. There is no clear indication to me that an elliptical pattern is any advantage. For example, the woofer is not and cannot be made to be elliptical so this makes for a bad power response transition at the crossover. That is likely to be more audible than any supposed reduction in floor or ceiling bounce (which according to Griesinger are not audible). And the narrower vertical dimension means a less than attractive lower bound on the vertical directivity control, unless the vertical width is held at the same as the round version, in which case the width becomes excessive. For a given baffle area availability I still contend that round is the better compromise.
I think that people just want something different. They refuse to accept that what exists might be the best approach. Especially if its not their idea.
Last edited:
The De250 looses it completely above about 16-18 kHz. Is this "controlled?" I don't think so. There are sometime breakups evident at frequencies below that, but nothing pronounced. In the critical range from 1 kHz - 5 kHz it is a very well behaved unit. Not so some other 1" drivers that I have tested.
@gedlee:
Thanks, that is all I need to know.
@weltersys:
Nice to learn about first hand experience with the Hughes transition.
@WithTarragon:
Do you know of any free source for the AES presentation paper?
Thanks, that is all I need to know.
@weltersys:
Nice to learn about first hand experience with the Hughes transition.
@WithTarragon:
Do you know of any free source for the AES presentation paper?
Okay lets try with pictures:
that (radial) :
versus this (same profile, but not radial) :
