I concur, too close is misleading, especially for polars. I find that 1.5 m from the mouth is a good compromise.
Polars can show resonances and diffraction as seperate issues. Single axis cannot differentiate them.
OS look like conical because they asymptomatic approach conical at large distances from the throat. But near the throat they are entirely different and small differences have huge effects at this location.
I don't see it as a random easement. The problem is that a planar emanation has pressure gradients that run contrary to the desired wavefront shape. This needs to be managed as soon as possible. The OS shape uses the available tools, such as diffraction, to pull it into line to significantly reduce the entrainment of these modes.
Do you agree that the best waveguide is based around a cone? A room is an odd shape (to sound)..
The LeCleach horn is a beautiful horn but it is intended to terminate into free space, making working with a room a compromise. It also has a falling directivity index.
The Hyp/Ex is great for a lower range horn because its diffraction in the middle gets good directivity control for a reasonable size, and the issues are beyond perception. It can be terminated into half space, but for the wrong reasons.
However only a cone is not meant to be terminated. It is terminated by force when there is no practical alternative. It is the correct beginning for a wavefront that will fulfill two important goals that I have.. One is to work with the straight walls of a room for the longest time possible, and the other is to fill as much of the room with sound before the complications of reflections take over.
I don't see the curved throat as having much effect below the higher frequencies, as witnessed by a plot of the acoustical resistance of an OS waveguide. A more conventionally curved horn is going to confine higher frequencies, requiring them to diffract at a later time, making the horn longer, as well as deviating from constant directivity, only to afford greater loading for the lower frequencies.
It's all about degree. One can use a simple radius to a pure cone - which is then not OS - and it works "almost" as well because its is "almost" OS. OS can be proven to be the lowest diffraction possible in a flared device. The closer you get to OS the lower the diffraction. If we know what the optimum is, then why would we use anything else?
I'm not sure that the acoustical resistance is any proof of anything, but yes, if the wavelengths are very long then the sound waves don't even see the shape. They see the throat area and the mouth area, anything in between is not important at LF. But once the wavelengths get shorter than the cross section then the details of the shape become important - very important.
AN OS waveguide can hold a directivity much wider than the equivalent circular piston of its throat (contrary to popular belief). But deviate from the OS even slightly and this control will fail earlier. The more you deviate the lower the frequency of failure.
The wave-front at the entrance to a waveguide is indeed a factor, but remember that it is the intention of the phase plug to create a flat wave-front at the throat. Is this achieved? Most likely not, but the extent to which its imperfections are an issue is not obvious. Much evidence (my own measurements) indicate that there is not a problem with the wave-front at the throat of a properly designed waveguide because there are not significant polar aberrations even at the highest frequencies.
We looked at this issue at B&C in some detail, but as far as I can tell there is nothing to be gained by improving the phase plug to be more coherent (flatter), despite the fact that I have a patent on how to achieve such an improvement.
Hence, while we can argue that this is theoretically a problem, there is no evidence that it is.
We looked at this issue at B&C in some detail, but as far as I can tell there is nothing to be gained by improving the phase plug to be more coherent (flatter), despite the fact that I have a patent on how to achieve such an improvement.
Hence, while we can argue that this is theoretically a problem, there is no evidence that it is.
And wouldn't measurements of different throat profiles and different drivers show you what works and how well, anyway? Meaning that while you may not know the exact shape of the wave front, acoustic measurements will tell you how well a given throat profile works and how well it conforms to predictions.
I think this measurement issue has been discussed to death. The reason why I recommend at measuring from inside the horn to out is to look at how the frequency response shifts as you move out. I would strive for no shifts in the response if I went back to working on horns. Right now I am in Guangzhou ready to visit the Loudspeaker Sourcing show later today.
Maybe it has been "discussed to death", but if you think that "measuring from inside the horn to out is to look at how the frequency response shifts as you move out." is an important criteria then it hasn't been discussed enough. I do not see what importance such a measurement would have. The near field of any source changes with location, but that's irrelevant. It's what gets propagated to the far field that matters, not the near field.
With 3D printing so convenient today, Earl’s phase plug concept could easily be printed and tested. It probably would cost $100 to get a metal one made of brass or copper.
How much difference will it make? Compared to OS vs a simple round throat? We may never know until we test. My patents are mostly tested to verify improvements before actually applied for.
How much difference will it make? Compared to OS vs a simple round throat? We may never know until we test. My patents are mostly tested to verify improvements before actually applied for.
Last edited:
Shifts in measurements of direct radiating drivers are due to diffraction from non-ideal transition of surface from cone to baffle. A horn should not have this problem with perfect flat wave front and uniform velocity with an OS throat is what I perceive. So if there is not much difference, then do we really need an OS throat? Plane wave to match it? Why go through all the fuss anyway? This is where I would actually listen to see if I can hear a meaningful difference which makes sense to carry the research to the next step.
All sources of sound have near-field and far-field effects, which is nothing more than saying they have "non-propagating effects (near-field,) and propagating effects (far-field.) Even perfect OS waveguides have both effects and the always imperfect mouth termination will also have both effects. Unless you are well away from the waveguides mouth you will measure both effects to various extents.
The "fuss" is about "optimization". The OS is the best that you can do - its "optimum." The issue that does not seem to matter is the flatness of the throat wave-front. The waveguide contour matters a lot.
I will always place my bet on the measurements and not any individuals uncontrolled listening test. If I can measure a difference then there is something that perhaps someone could hear. But if I can't measure a difference then there is no difference to be heard (not that people won't believe that they hear it anyways.)
The "fuss" is about "optimization". The OS is the best that you can do - its "optimum." The issue that does not seem to matter is the flatness of the throat wave-front. The waveguide contour matters a lot.
I will always place my bet on the measurements and not any individuals uncontrolled listening test. If I can measure a difference then there is something that perhaps someone could hear. But if I can't measure a difference then there is no difference to be heard (not that people won't believe that they hear it anyways.)