A new approach to point source?

[Heinrich]

So the trick imo, in any of the ring/array type speakers that use mids around a central tweeter, is to first have a tweeter that reaches low enough to allow a c2c between the mids to be within 1/4 WL.

O/k, if there is no ripple due to sufficiently small c2c distance, the problem with vertical lobing would not exist in the first place. So there would be no need for a solution either. All the effort literally for nothing.

That‘s why I asked the board to state the problem to be solved, but didn‘t get any answer yet. One may discuss and praise fanta solutions, but without the problem, what to say?

In DIY such would be utterly o/k, experiment, gain experience, approach unknown grounds. But then a final, comprehensible report is missing all too often. (I asked, no answer.)

In the commercial market such cannot be recommended for the cost of risk alone.

 

[Audionaut]

When I had an extended listening session, sound quality in reality could not keep up, to say the least.

Interesting, I guess I'll have to hear for myself, but the options are very limited here.

I haven't heard the Beolab yet, but you're the first to dislike it.
Have you been able to adjust them? Where did you hear them?

 

[Juhazi]

Some very good posts and examples above...

Omnidirectional (dodecahedral) speakers are used in acoustic studies, but their bandwidth is limited typically to 250-2500Hz. For home hifi there are many more or less potent solutions from eg. Bose, Sonab, MBL and B&O

Point source means similar directivity horizontally and vertically, without definition of DI or -6dB angle (radiation pattern) With modern knowledge of acoustics and transducers, there are many ways to control directivity. https://www.klippel.de/fileadmin/_migrated/content_uploads/KLIPPEL_Sound_Radiation_Poster_01.pdf

But there is no ideal single best solution, it depends on for what purpose the speaker is used and the environment.

 

[thewas]

That‘s why I asked the board to state the problem to be solved, but didn‘t get any answer yet.

One answer given already few times in this thread is asymmetric horizontal (wide) and vertical (narrow) directivity which isn't possible with coaxial drivers.

 

[gnarly]

Exactly, this would be also my personal way to differ between point source and not.

Thanks for that comment. I was hoping a pragmatic definition of 'point source' would make sense to some.

In the real world, some additional aspects factor in, like diaphragm geometry and resulting directivity, but as a rule of thumb this sounds very reasonable.

Yes, I think the name of the game is to consider those additional aspects up front, and then stich everything together within 1/4 WL spacing.
A MEH/unity/synergy is the best route for this i've found so far.

The main problem with these circular arrays following such strict rules is the tweeter. There is basically no way to have a tweeter array, so every midrange or midwoofer circle around it must somehow adopt to the directivity behavior of the single tweeter itself. But which one is the best?

Yes again. It's almost impossible to get a home audio tweeter strong enough, and able to reach low enough, to make a circular array work within 1/4 WL spacing.
A compression driver on a small shallow waveguide, one that can reach down to 5-700Hz, is about the only way I know, that a circular array might work.
And I view tweeter arrays as simply a b-a-a-d idea.

O/k, if there is no ripple due to sufficiently small c2c distance, the problem with vertical lobing would not exist in the first place. So there would be no need for a solution either. All the effort literally for nothing.

I need to say I'm either not following you.......or .....are you kidding me !?
Virtually no speaker achieves the small c2c distances necessary to avoid ripple...
It's a very real problem, with a continual search for a solution.

 

[kharan]

One answer given already few times in this thread is asymmetric horizontal (wide) and vertical (narrow) directivity which isn't possible with coaxial drivers.

There are non-circular drivers, and Scan Speak makes a non-circular motor structure too. The idea of a coaxial driver that is non-circular and would provide an different vertical and horizontal directivity is plausible.

There are probably reasons why this isn't done, but calling it impossible is premature imho.

 

[thewas]

There are non-circular drivers, and Scan Speak makes a non-circular motor structure too. The idea of a coaxial driver that is non-circular and would provide an different vertical and horizontal directivity is plausible.

There are probably reasons why this isn't done, but calling it impossible is premature imho.

I should have made it more clear I was talking about coaxial drivers currently existing in the market. The non circular Ellipticor Scan Speak drivers show how difficult it is to make such a normal driver, not to talk about a coaxial one, but of course it is not impossible. Also their dimension asymmetry is quite limited, so not for big directivity differences like in the DIY array I posted above.

 

[Heinrich]

When drivers' c2c's can be kept within 1/4 wavelength throughout their frequency summation ranges, they can radiate polars without ripple.
So the trick imo, in any of the ring/array type speakers that use mids around a central tweeter,

O/k, if there is no ripple due to sufficiently small c2c distance, the problem with vertical lobing would not exist in the first place. So there would be no need for a solution either. All the effort literally for nothing.

I need to say I'm either not following you.......or .....are you kidding me !?
Virtually no speaker achieves the small c2c distances necessary to avoid ripple...
It's a very real problem, with a continual search for a solution.

You really got me plain right.

That‘s why I asked the board to state the problem to be solved, but didn‘t get any answer yet.

One answer given already few times in this thread is asymmetric horizontal (wide) and vertical (narrow) directivity which isn't possible with coaxial drivers.

The "solution" leads to a vertical beamwidth of +/-15°, as exemplified in post #8. And up to about 250Hz the bass is designed to be virtually omnidirectional.
Still the question is not answered, by what problem the designer might be pushed into this pretty narrow (!) corner. At home, unusable, plain and clear.

What is this tinkering after?!

 

[thewas]

The "solution" leads to a vertical beamwidth of +/-15°, as exemplified in post #8. And up to about 250Hz the bass is designed to be virtually omnidirectional.
Still the question is not answered, by what problem the designer might be pushed into this pretty narrow (!) corner. At home, unusable, plain and clear.

What is this tinkering after?!

Such narrow vertical radiation can have benefits at typical home high listening distances and reflectivity. Directivity in the modal region is less a problem due to various reasons.

 

[Heinrich]

Such narrow vertical radiation can have benefits at typical home high listening distances and reflectivity. Directivity in the modal region is less a problem due to various reasons.

C'mon.

 

[gnarly]

You really got me plain right.



The "solution" leads to a vertical beamwidth of +/-15°, as exemplified in post #8. And up to about 250Hz the bass is designed to be virtually omnidirectional.
Still the question is not answered, by what problem the designer might be pushed into this pretty narrow (!) corner. At home, unusable, plain and clear.

What is this tinkering after?!

Sorry I have no clue what you are talking about. But that's OK.. doesn't really seem worth our pursuing...

 

[Arindal]

The idea of a coaxial driver that is non-circular and would provide an different vertical and horizontal directivity is plausible.

You take the words right out of my mouth! I am working on exactly such a solution, as a side project of achieving adjustable yet constant directivity index over a broad band.

There are probably reasons why this isn't done, but calling it impossible is premature imho.

It is anything but trivial, taking into account geometry, directivity properties, potential x-over freq and subjective sound quality of available drivers. At the moment it look like it is doable, with a certain probability.

Have you been able to adjust them? Where did you hear them?

In a B&O store with a suitable listening room. And yes, every adjustment was possible and there was someone from the store who knew what he was doing. It is not that these are bad speakers, and the beamsteering works astonishingly well. It is just they sound like any tweeter array in the treble. You notice it particularly when listening to complex material demanding an optimum of subjective transparency in the treble region (orchestra with massive brass, choirs, big band and alike).

Point source means similar directivity horizontally and vertically,

Interestingly, you introduce another, different definition which I have never heard of. Would rather call this ´rotationally symmetrical directivity´.

 

[PristineSound]

You take the words right out of my mouth! I am working on exactly such a solution, as a side project of achieving adjustable yet constant directivity index over a broad band.

Like an oval shaped driver coaxial? I can see that being plausible. But I have never, ever seen one, if you can design one (not counting car audio, which is suspect), it would be the first.


Edit: this one seems less suspect

 

[Arindal]

Like an oval shaped driver coaxial?

Neither oval nor conventionally coaxial in the sense of tweeter in the center of the midrange´s voicecoil. I am not a driver designer.

Edit: this one seems less suspect

Interesting design, but I would not follow that route even if I could. From theoretical point of view, I would not want a driver to combine broad horizontal tweeter dispersion with narrow horizontal midrange pattern and vice versa. That is what you basically get when using an oval shaped, more or less steep cone as the waveguide of your tweeter.

 

[thewas]

C'mon.

?? Please elaborate.

 

[Heinrich]

Like an oval shaped driver coaxial? I can see that being plausible.

... plus all the following from the DIY department.

The idea that a 'point source' can be represented by means of a 'mid array' has already been exposed as a fallacy. Neither is the concept of a 'pont source' in any way conclusive, nor does the 'mid array' generally solve the lobing problem from the individual driver distances in a multi-way box.

This answers your original question (post #1).

Now the topic is brought to a narrow radiation in the vertical. But I ask again, for what purpose? A design was advertised that restricts vertical directivity to a synmmetric +/-15°. The radiation beyond +/-15° is also sharply cut off, it does not gradually attenuate. In the horizontal plane, however, the box should radiate very broadly. It would therefore be more of a line source than the circular thing you asked about.

When asked for what purpose this is done, the only answer is again

Such narrow vertical radiation can have benefits at typical home high listening distances and reflectivity. Directivity in the modal region is less a problem due to various reasons.

But what are the claimed benefits? It is not clear in any way. I would like to point out that the bass and lower mid-range reproduction with that box is omni-directional. A huge break in the directivity pattern. It is claimed that there is a connection with the Schröder frequency. But it is not explained what this connection is. And I suspect that there is none.

Now you come along with a car loudspeaker to ask the contributors to evaluate this construction method according to their ideal conceptions.

What is it? A decided yes and no, but ... .

Look, the industry has a lot of solutions to offer, but what is the problem? I leave it at that, as I'm expressively ignored, which I take personal.

Good luck

 

[Blockader]

... plus all the following from the DIY department.

The idea that a 'point source' can be represented by means of a 'mid array' has already been exposed as a fallacy. Neither is the concept of a 'pont source' in any way conclusive, nor does the 'mid array' generally solve the lobing problem from the individual driver distances in a multi-way box.

This answers your original question (post #1).

Now the topic is brought to a narrow radiation in the vertical. But I ask again, for what purpose? A design was advertised that restricts vertical directivity to a synmmetric +/-15°. The radiation beyond +/-15° is also sharply cut off, it does not gradually attenuate. In the horizontal plane, however, the box should radiate very broadly. It would therefore be more of a line source than the circular thing you asked about.

When asked for what purpose this is done, the only answer is again

But what are the claimed benefits? It is not clear in any way. I would like to point out that the bass and lower mid-range reproduction with that box is omni-directional. A huge break in the directivity pattern. It is claimed that there is a connection with the Schröder frequency. But it is not explained what this connection is. And I suspect that there is none.

Now you come along with a car loudspeaker to ask the contributors to evaluate this construction method according to their ideal conceptions.

What is it? A decided yes and no, but ... .

Look, the industry has a lot of solutions to offer, but what is the problem? I leave it at that, as I'm expressively ignored, which I take personal.

Good luck

There’s no solid evidence that ceiling reflections offer any real benefit. Their interaural cross-correlation is low, and all they do is to introduce destructive interference with the direct sound, reducing both clarity and harming timbre. Floor reflections can be even more destructive, yet Toole argues they’re essential for realism because our ears expect to hear them. Although I appreciate the logic, my own listening experiences don’t support his claim.

In short, speakers with narrower vertical beamwidth deliver greater clarity because they are subject to less SBIR from ceiling and floor reflections. That said, vertical directivity still plays a very important role in a speaker’s sound power output, which affects far-field performance and at very long listening distances, sound-power becomes the primary determinant of perceived sound quality.

Ideal speakers for me would be something with very narrow vertical directivity and a wide controlled horizontal directivity.

 

[Heinrich]

There’s no solid evidence that ceiling reflections offer any real benefit.

It isn't bad either. I designed a mini speaker with deliberately wide dispersion. It works, listening to the playback is a joy. If you don't believe me, look for Grimm's LS1, which is praised for wide directivity in all (!) directions. Where is the problem?

Their interaural cross-correlation is low, and all they do is to introduce destructive interference with the direct sound, reducing both clarity and harming timbre.
Floor reflections can be even more destructive, yet Toole argues they’re essential for realism because our ears expect to hear them. Although I appreciate the logic, my own listening experiences don’t support his claim.

Cross correlation in stereo is a mess. It all relies on a person sitting just in the middle at the right hight not rotating the head a bit and staying there forever. All such technical demand and description breaks apart, once the gear is used for what it is made for: real people in casual spare time mode at home. We might, once in time, depart from dealing with stereo as if it were the "therapy" in Clockwork Orange (the movie).

In short, speakers with narrower vertical beamwidth deliver greater clarity because they are subject to less SBIR from ceiling and floor reflections. That said, vertical directivity still plays a very important role in a speaker’s sound power output, which affects far-field performance and at very long listening distances, sound-power becomes the primary determinant of perceived sound quality.

Ideal speakers ...

Did you ever hear of flutter echo? I leave it to you, your quest, my final statement.

 

[Blockader]

Did you ever hear of flutter echo? I leave it to you, your quest, my final statement.

Was that supposed to be your grand mic-drop moment where we were meant to be dazzled by how clever you think your answer was?
Even in a moderately furnished room with a few carpets, flutter echo is not an issue.

Returning to the IACC point you raised: when I’m at home and listening through my speakers, I always sit in the sweet spot. I mostly listen to music with headphones only switch to speakers when I really want to enjoy music. The point still stands though vertical reflections offer no benefit.

 

[Arindal]

The idea that a 'point source' can be represented by means of a 'mid array' has already been exposed as a fallacy.

By whom?

nor does the 'mid array' generally solve the lobing problem from the individual driver distances in a multi-way box

Do not see any evidence for that as long as the distance between the individual drivers is sufficiently small to prevent lobing. It is maybe not as easy to achieve with a ring of midrange cones around the tweeter being crossed over at a higher frequency, but it seems doable.

I designed a mini speaker with deliberately wide dispersion. It works, listening to the playback is a joy.

Try it in a reverberant environment in 4m of listening distance with complex sacred music. You will hear the limitations coming with a dominant indirect soundfield. And I personally would not call this a joy, regardless how good the speaker is. Can fully confirm everything member @Blockader has experienced.

Cross correlation in stereo is a mess. It all relies on a person sitting just in the middle at the right hight not rotating the head a bit and staying there forever.

For absolutely perfect phantom localization stability, high interaural correlation might be necessary in stereo. For just having the impression that there is an imaging and ambience between the speakers, it works pretty well even if the listening position is far off the center. @Blockader ist 100% right on this topic.

Did you ever hear of flutter echo?

Flutter echoes are not appearing in smaller or medium sized rooms unless there is a very particular geometry, causing a single discrete, repeating, high initial time delay reflection via multiple reflective walls (for example tilted ceiling or barrel vault shaped ceiling). It is pretty likely that such a room has other problems as well which can be attenuated by a narrow vertical dispersion.