Geddes on Waveguides

I would say yes, because the 45 degree can avoid the side walls and the 60 degree cannot. This is not a trivial distinction. In terms of sound power, the larger the angle the greater the total encompased area - the area grows with angle as
r*sin(theta)*d(theta). Thats one of the main points about a "hole" on axis. In terms of sound power it is miniscule. This is an aspect of the visual presentation of polar maps that has always bothered me. The angular axis, to be proportional to power, should not be linear, but should go as arcsin(theta). This makes the difference between 60 and 45 quite large in fact.
 
The angular axis, to be proportional to power, should not be linear, but should go as arcsin(theta).

That may be backwards, but the point is that the angles should move slowly near the axis and much faster as they approach 90 degrees, to account for the much larger area represented by the data at larger angles. All angles are not equal in terms of power. This can also be seen in the calculation of DI where the data points are weighted by sin(theta).
 
Administrator
Joined 2004
Paid Member
I see what you're getting at and like the idea. I'm just not sure how to visualize it meaningfully on the charts. What is a good visual representation of the power response as the listener will actually hear it?

Once the speaker gets into a room with walls, floor and ceiling it's going to really change things. The CARA software does a decent job of showing this, but has its limitations.

Interesting line of thought. Hmmmm.....
 
In simplest terms, constant
By the way, ARTA can produce the type of sonograms (contour plots) {as well as surface plots (water falls) and polar plots} that have been presented here. Of course, the ATAR measurement system cost about $100.

Hello John,

I have Arta also, which is what I currently use to make a polars. But Arta can sometimes be quite complex for an occasional user like me... On the other hand holm is just so easy to work with.The feature in holm where you can adjust the gating and see the effect immediately in the frequency response was very educational...

I would have no problem spending 100$ on holmimpulse, if development would continue. In an ideal world the features of holm and Arta would be combined, and integrated in the holm user interface.
 
Here's what the scale would look like on a sonogram based on sin(theta). Kinda like log scaling but more compressed at the top end.

attachment.php
 

Attachments

  • sinscale.png
    sinscale.png
    620 bytes · Views: 491
John, I think the point is that my usage has always been consistant with the defacto standard used in the Pro industry. Why is it that we have to redefine all of our terms for use in DIY?

I don't know. Perhaps it is because most DIY'er are pros? In may case it is because I am going by the definitions of directivitive typically give in an acoustic text. Those same texts seem to define beam width as the angular distance between two point on either side of the principal axis where the response is down by 6dB.

I guess any professional filed is full of jargon which others don't necessarily share.
 
Other than as a mathematical abstraction, is the practical difference in a real room really that large between:

a. Earl's waveguides: -6dB at 45 degrees
b. A pure dipole: -6dB at 60 degrees?

Let's ignore low frequencies and the rear radiation pattern for now.

In a room with well damped walls not really. In a live end/dead end room like the experiments of the later 70's early 80's there won't be much pratical difference. In a more lively room, definately. This is what gives the dipole the characterist of openness and a sence of ambiance. I have always said that direct radiator speakers, and those like Earl's designed in paticualr, provide the listener with the greatest sense of what is on the recording. A "you are there" kind of effect. Dipole reporduction in a more lively room creates more of a "they are here" effect.

Now much detail you hear is really more dependent on how far you are form the source that anything else.
 
I submit "Defined Directivity."

JBL (Keele) used it to describe horns which varied with angle according to a prescribed pattern for a specific purpose and application, but despite "constant" being unattainable in the strictest sense, whether it is or not, or to what degree, "control" is clearly implicit. In his original paper describing constant directivity, Keele illustrated horn designs having varying beamwidths; beamwidth per se is not definitive, and must be independently specified.

We get into considerable difficulty when we conceptually expand application of these terms to systems as opposed to elements thereof. A system having varying wide dispersion at low frequencies and quasi-constant directivity of narrower beamwidth at high frequencies is a controlled directivity system only with respect a portion of its total bandwidth. By the same token, a dipole system having directivity control at low and mid frequencies but collapsing directivity above 10 kHz is not a constant directivity system either.

Both incorporate defined directivity, however, in the larger view, there is far more to be accomplished in exploring the merits of differing specific definitions with respect to application and performance than in haggling over the terminology used to describe them.... ;)
 

Attachments

  • Nathan Directivity per Geddes A.jpg
    Nathan Directivity per Geddes A.jpg
    67.4 KB · Views: 380
  • Orion Directivity per Geddes A.jpg
    Orion Directivity per Geddes A.jpg
    65.7 KB · Views: 459
  • Twoway Directivity per Geddes A.jpg
    Twoway Directivity per Geddes A.jpg
    67.8 KB · Views: 278
Last edited:
But this ignores:

1. the far side wall, and
2. the amount of additional listening distance from the loudspeakers (..if any).
And neither really 'avoids' the near side wall. The waveguide is -6dB and the dipole is -3dB at 45 degrees. A difference to be sure but is 3dB difference really night and day, especially when most people will put an absorber on the near wall? I suspect the difference in sound has more to do with the rear pattern than with different directivity up front.
 
A difference to be sure but is 3dB difference really night and day, especially when most people will put an absorber on the near wall? I suspect the difference in sound has more to do with the rear pattern than with different directivity up front.
Put an absorber on the near wall and you also kill the low-IACC contralateral first reflection providing spaciousness cues for the narrower dispersion design.

It is well recognized that imaging and spaciousness are in degree mutually exclusive with conventional approaches. With narrower-dispersion controlled-directivity designs, the listener may vary their balance by adjusting toe-in and listening distance according to their individual preference and pleasure.... :)

Edit:
Look at the angles involved, according to Linkwitz; the near-wall reflection comes from almost 90° off-axis:
 

Attachments

  • Linkwitz slide11c.jpg
    Linkwitz slide11c.jpg
    71.4 KB · Views: 282
  • Linkwitz slide11d.jpg
    Linkwitz slide11d.jpg
    51.5 KB · Views: 282
Last edited:
Put an absorber on the near wall and you also kill the low-IACC ipsilateral first reflection providing spaciousness cues for the narrower dispersion design.

It is well recognized that imaging and spaciousness are in degree mutually exclusive with conventional approaches. With narrower-dispersion controlled-directivity designs, the listener may vary their balance by adjusting toe-in and listening distance according to their individual preference and pleasure.... :)
Toole's book, as I understand it, says first reflections from the near wall are bad (correlated and too soon behind the the direct sound) but first reflections from the far wall are good (delayed enough and uncorrelated with the direct sound to the opposite ear.)
 
Last edited:
Yes Zilch, I think we're in vigorous agreement here. :) A dipole set up like Linkwitz's diagram wouldn't need an absorber because of the 90 degree null. But if you had a speaker closer to the wall and maybe not at that angle, you'd probably want to absorb the near wall reflection but not the far wall reflection (small absorber).
 
The far wall is just that, far. The SPL drops with the square of the distance.
Yes, but that reflection is also generated more on-axis. According to Toole Fig. 8.5, discounting directivity, -2 dB for the ipsilateral reflection, and -6.3 dB for the contralateral. It's the delays that most matter with respect to their relative detrimental influence, 2.7 ms and 12.3 ms, respectively.

With on-axis toe-in, the ipsilateral comes from -73° (Fig. 16.6) in Toole's standard room and setup. From Earl's measurements, the dipole does not appear to be as effective as the waveguide in suppressing the ipsilateral reflection at that angle in the frequency range of primary interest:

http://www.diyaudio.com/forums/multi-way/103872-geddes-waveguides-453.html#post2235918
 
Last edited: