Why are there so few omnidirectional speakers?

[Duke]

I do wonder, what the impact would be within a more rear space constrained setup (ie: closer to the wall behind it) of reversing the delay, and configuring them for a simulated point source 30 to 35 cm in front of the speaker ...

Offhand I don't see how one could configure the Quads, or any other speaker, to simulate a point source that was closer to you. Maybe I'm missing something obvious?

Obviously not for a near field listening setup - but in a standard setup, this might allow for placing the speakers within circa 50 to 80cm of the rear wall

The time delay between the direct sound and the backwave reflection is a function of the path length difference, not the delay line function. Everywhere that the frontwave output is delayed, the backwave is correspondingly delayed, so the path length difference is still the determining factor.

Not saying there are no work-arounds for dipole owners who have severe space constraints, but imo that would not be among them. Unless I'm missing something... which could well be the case.

In some of my multi-directional designs I deliberately aim the "backwave" energy in a direction which results in a fairly long reflection path length, specifically to allow placement fairly close to the wall behind the speakers. Perhaps a reflector could be designed which would do the same thing for a dipole speaker? One limitation would be that long wavelengths could just wrap around the reflector.

 

[dlaloum]

Offhand I don't see how one could configure the Quads, or any other speaker, to simulate a point source that was closer to you. Maybe I'm missing something obvious?

The ESL63 and subsequent models uses 2 elements (for the midrange) with 7 circular concentric conductors, which are fed by an all pass network that acts as an electrical delay line. This elegant construction gives the ESL both the character of a point source and an accurate time relation between high and low frequencies.

(quoted from: https://www.meddens.eu/audio/esl63.htm )

The key thing is by using circular concentric delay system, it simulates a point source - with the central area being early, and as you move out the sound is ever more delayed, as it would be if you were hearing a point source about 30cm behind the speaker.

The interesting thing in this case - is that you can reverse the delay system (should you choose to) and make it simulate a point source 30cm ahead of the speaker - I am just curious if you then think of the system as a point source 30cm in front of the speaker location, can you then shift the speaker closer to the wall... have you effectively gained 30cm... (or rather - you have gained the delay equivalent to 30cm... which is all that is of import in this case, when it comes to reflections and perception)

 

[Newman]

It simulates a four-inch point source….including the beaming.

(Actually the 989 variant)

 

[gnarly]

A few random thoughts, hopefully fitting in...

I think dipoles and omnis are different enough in their radiation patterns that it's hard to compare them.
As dipoles typically have much less vertical radiation.

The smaller the room, the more heavily damped the room, and the closer the listener is to the speakers, are all factors that tend to increase the ratio of direct to reflected sound. Which makes omnis and dipoles sound closer to each other, or even closer to point sources.

That said, I've never been in small enough, damped enough room, to not be able to know i was listening to an omni.
Whereas I can't say the same thing about being able to tell a dipole vs point source.

In the end, i think it all comes down to the ratio of direct vs reflected sound, and the time delay (or longer reverberations) of the reflected.

 

[Duke]

The ESL63 and subsequent models uses 2 elements (for the midrange) with 7 circular concentric conductors, which are fed by an all pass network that acts as an electrical delay line. This elegant construction gives the ESL both the character of a point source and an accurate time relation between high and low frequencies.

(quoted from: https://www.meddens.eu/audio/esl63.htm )

The key thing is by using circular concentric delay system, it simulates a point source - with the central area being early, and as you move out the sound is ever more delayed, as it would be if you were hearing a point source about 30cm behind the speaker.

The interesting thing in this case - is that you can reverse the delay system (should you choose to) and make it simulate a point source 30cm ahead of the speaker - I am just curious if you then think of the system as a point source 30cm in front of the speaker location, can you then shift the speaker closer to the wall... have you effectively gained 30cm... (or rather - you have gained the delay equivalent to 30cm... which is all that is of import in this case, when it comes to reflections and perception)

I owned a pair so I'm somewhat familiar with how it behaves, though I don't know the circuit they used to get the delay.

I don't see how reversing the delay would "make it simulate a point source 30cm ahead of the speaker". What you'd have is the outer rings radiating first, then the next ones, and so on until the middle. The resulting pattern might kinda sorta simulate that of a really big and deep cone, but I don't think it would simulate a point source in front of the panel because the arrival times would be wrong.

 

[Newman]

…and of course a true point source wouldn’t have anything like the radiation pattern I showed above.

What Walker did achieve was a mechanism to prevent the horrible on-axis panel cancellations that would have occurred without the delay lines. Although these are less audible as distance from the panel increases. I wonder whether Walker was especially conscious of relatively small English listening rooms, and went to a lot of trouble do deal with near field on-axis cancellations only. To an ear in that position, the Quad’s direct sound is more like a point source and less like a large flat panel. But to say that it radiates like a point source, implies a whole lot of off-axis niceness that the Quad lacks.

cheers