Why aren't there more very wide directivity speakers?

[Mort]

I remember reading that Toole reported a possible preference for wide (controlled) directivity over narrower.

Even if that's not true, why do we see fewer wide directivity speakers? Are they harder to design? More expensive?

 

[jeffaegrim]

Isn’t it the other way around? I thought that the majority of speakers circulating around are in fact cone/dome types which will radiate sound in a wide manner (though maybe not necessarily controlled). A platonic form of a modern speaker is something small with a dome tweeter and a woofer, I feel

And it’s narrow DI speakers that one sees rarely - mainly because they tend to be pretty big

 

[Mort]

Isn’t it the other way around? I thought that the majority of speakers circulating around are in fact cone/dome types which will radiate sound in a wide manner (though maybe not necessarily controlled). A platonic form of a modern speaker is something small with a dome tweeter and a woofer, I feel

And it’s narrow DI speakers that one sees rarely - mainly because they tend to be pretty big

Maybe, you make a good point.

I was thinking 'wide' starts at 70.
50-60 is 'medium' and where I see most domes. 'Narrow' being below 50.

I don't see most being over 70 but my arbitrary definition might be off.

 

[Jim Taylor]

... why do we see fewer wide directivity speakers? Are they harder to design?

There are two factors at work here. One is money. The tremendous investment and research into cone (and dome) drivers makes them hands-down the major market choice. No panel speaker can match their consistency and refinement in the search for bandwidth and output per dollar.

The second is a consequence of that choice. Extremely wide dispersion speakers would take lateral arrays of multiple "normal" cone/dome drivers, and that means interference and comb filtering.

Any entity investing in radical new technology to sidestep the above had better have a great deal of time as well as cash reserves ... and that debt will be an albatross around the neck of the company.

TANSTAAFL

 

[jeffaegrim]

Maybe, you make a good point.

I was thinking 'wide' starts at 70.
50-60 is 'medium' and where I see most domes. 'Narrow' being below 50.

I don't see most being over 70 but my arbitrary definition might be off.

Definitely agree if you put it this way, and thinking of domes as „medium“ wide probably makes more sense. I normally consider a 9-10db+ DI from ~1khz upwards *narrow*

I have a pair of very wide and very narrow radiating speakers in my living room right now and making direct comparisons has been incredibly enlightening and interesting. Regarding the latter, they are the JBL CBT70 measured here and they are absolutely phenomenal - without suffering from some FR irregularities that Amir found, which must be particular to the sample that was sent in. I still need to report my findings in that thread, as I became a big fan of those speakers. Soundstage wide as a house with nigh infinite dynamics

 

[Duke]

I remember reading that Toole reported a possible preference for wide (controlled) directivity over narrower.

Even if that's not true, why do we see fewer wide directivity speakers? Are they harder to design? More expensive?

Interesting question!

Would you consider the Revel Salon 2 to be a "wide directivity" speaker? Soundstage's measurements include horizontal off-axis data out to 90 degrees in 15-degree increments; one might argue that its radiation pattern is pretty wide up to about 10 kHz:

SoundStage! Measurements - Revel Ultima Salon2 Loudspeakers (12/2009)

Achieving a wider horizontal pattern than the Salon 2's 1" dome tweeter well north of 10 kHz calls for some juggling of tradeoffs, as @Jim Taylor mentioned... either an array of some kind, or a narrow ribbon (whose vertical dispersion will be reduced due to its height), or maybe a really tiny dome supertweeter, or (my preference) perhaps adding a rear-firing tweeter to fill in the missing off-axis energy in that top octave or so.

Is there strong evidence that even wider horizontal directivity than something like the Salon 2 would be preferred? The Salon 1 had a rear-firing tweeter, but that was not retained in the Salon 2, almost certainly based on listening tests.

Or, is the preference trend evidently in the direction of even wider directivity (even if it's not yet "proven" to be preferred), and therefore you're reasoning that wider directivity - but presumably still less than full-on omnidirectional - might be worth pursuing?

As to whether very wide directivity speakers are "harder to design? More expensive?"... I would say yes to both, but not prohibitively so.

I like the idea of exploring alternative approaches for which there is indication of validity in the data we already have access to, whether or not said alternative approaches are yet "proven".

 

[Mort]

I have a pair of very wide and very narrow radiating speakers in my living room right now and making direct comparisons has been incredibly enlightening and interesting.

Yes! Last year I got to A/B a pair of coaxial Genelec domes (~50 iirc) with a philharmonic audio HT tower at 80ish.

It was so revealing. It does affect the sound stage in interesting ways. I like both lol.

I want to try something narrower but that's even harder as you point out.

 

[Mort]

Would you consider the Revel Salon 2 to be a "wide directivity" speaker?

Yeah I would, anything over 70 that holds together counts for me.

As to whether very wide directivity speakers are "harder to design? More expensive?"... I would say yes to both, but not prohibitively so.

Great post. Thanks for that.

You raise good questions about my question and I'll acknowledge some limitations. One being I have not heard an omnidirectional speaker. I probably should. It's there a sweet spot? Maybe revel found it in the salon 2

 

[MKR]

Yes! Last year I got to A/B a pair of coaxial Genelec domes (~50 iirc) with a philharmonic audio HT tower at 80ish.

It was so revealing. It does affect the sound stage in interesting ways. I like both lol.

I want to try something narrower but that's even harder as you point out.

Great question … well, if you like both (as do I, depends upon the recording), then variable directivity is the answer! My speakers have a user selectable (via remote) narrow, wide, and omni mode. Best of all worlds.

 

[Mort]

Great question … well, if you like both (as do I, depends upon the recording), then variable directivity is the answer! My speakers have a user selectable (via remote) narrow, wide, and omni mode. Best of all worlds.

Fortunately I have two listening spaces!

What speakers have variable directivity? I had forgotten about that.

 

[MKR]

Fortunately I have two listening spaces!

What speakers have variable directivity? I had forgotten about that.

Soundfield Audio. Mine are custom, but they offer this capability with many of their designs. Also I believe the Beolab 90s offer variable directivity. I am sure there are others, but I am not aware of them.

 

[Arindal]

I would rather tend to ask the opposite question: Why are there so many wide-radiating speakers on the hi-fi market, and few narrow ones?

We should define what is wide and narrow, and how typical home listening conditions look like. If we take typical listening distances and reverberation time of a living room into account, and expand the question ´wide or narrow?´to all relevant frequency bands (i.e. lower midrange included), 95+ per cent of home speakers are too wide in directivity.

Toole reported a possible preference for wide (controlled) directivity over narrower.

With all due respect, but this hints to the ´Circle of confusion´ being at play. If you ask participants of listening tests solely for their subjective preference, with them being completely unfamiliar with recordings, listening room and channel layout, nor having a clear reference how recordings should sound like, well you end up them them choosing the least annoying variant. That is particularly the case with recordings not containing meaningful, natural reverb, like almost all pop, electronic and rock productions.

was thinking 'wide' starts at 70.
50-60 is 'medium' and where I see most domes. 'Narrow' being below 50.

You are referring to the listening window width horizontally? If you take the lower midrange into account, I would say 99% of home speakers can be defined as ´wide´ then.

 

[kemmler3D]

Purifi SPK16 is apparently ~70deg, and apparently it took serious effort in simulation and WG design to get it that way. I get the impression that it's pretty hard to get a tweeter to stay wide at high frequency, so the more sensible approach is to make it go narrower in lower frequencies to match a woofer.

 

[Arindal]

As to whether very wide directivity speakers are "harder to design? More expensive?"

If you opt for drivers being small compared to the wavelengths emitted, and a comparably slim baffle, a wide directivity speaker is actually pretty easy and cheap to build. Take a 6.5" woofer, a 2.5" midrange and 0.75" dome, choose low crossover frequencies - voilà!

Is this a desirable concept? For home listening, rather not. Wide dispersion usually comes with higher SPL being radiated towards ceiling, floor and side walls, resulting in dominant early reflections and degrading localization. For nearfield monitoring in a well-treated studio, it might be an excellent solution, though.

f you like both (as do I, depends upon the recording), then variable directivity is the answer!

That is really the one thing that is difficult to implement over a vast frequency range. Not impossible, though, as Duke has mentioned. An adjustable rear-firing tweeter (plus upper midrange if necessary), would do. Sonus Faber has done something like that in the past:

Such concepts are usually resembling adjustable bipoles in the treble bands, but their sound is heavily dependent on how reflective/diffusive the rear wall behind the speakers is. I guess that was the reason why SF have implemented a mechanical possibility to alter the horizontal angle of the rear-firing unit.

Note such are concepts to widen directivity hence reducing directivity index. Adjustable directivity concepts to narrow the beam, such as the Beo 90, are much more complicated to design and usually involve driver arrays, which I would recommend for tweeters.

 

[dlaloum]

Achieving a wider horizontal pattern than the Salon 2's 1" dome tweeter well north of 10 kHz calls for some juggling of tradeoffs, as @Jim Taylor mentioned... either an array of some kind, or a narrow ribbon (whose vertical dispersion will be reduced due to its height), or maybe a really tiny dome supertweeter, or (my preference) perhaps adding a rear-firing tweeter to fill in the missing off-axis energy in that top octave or so.

The Gallo CDT Tweeter appears to be a high quality wide directivity solution...

 

[ppataki]

We should define what is wide and narrow, and how typical home listening conditions look like.

Just checking my understanding on narrow and wide dispersion

A tweeter (SB26STWGC-4) with this off-axis response:

Would this considered to be wide or narrow?

At 10kHz 30 degrees approx. -2.5dB, at 60 degrees approx. -5dB
At 16kHz 30 degrees approx. -2dB, at 60 degrees approx. -5dB
At 20kHz 30 degrees approx. -6dB, at 60 degrees approx. -22dB

Claude told me this is considered a wide dispersion (and I would tend to agree with that) but would be nice to understand what is considered wide or narrow

 

[Arindal]

Would this considered to be wide or narrow?

Below 16K definitely wide. It is a bit difficult to judge if we do not have data from the rear hemisphere (90deg+) and no information on baffle/no baffle/waveguide situation.

 

[ppataki]

Below 16K definitely wide. It is a bit difficult to judge if we do not have data from the rear hemisphere (90deg+) and now information on baffle/no baffle/waveguide situation.

Thank you, this confirms my understanding too

 

[MKR]

If you opt for drivers being small compared to the wavelengths emitted, and a comparably slim baffle, a wide directivity speaker is actually pretty easy and cheap to build. Take a 6.5" woofer, a 2.5" midrange and 0.75" dome, choose low crossover frequencies - voilà!

Is this a desirable concept? For home listening, rather not. Wide dispersion usually comes with higher SPL being radiated towards ceiling, floor and side walls, resulting in dominant early reflections and degrading localization. For nearfield monitoring in a well-treated studio, it might be an excellent solution, though.



That is really the one thing that is difficult to implement over a vast frequency range. Not impossible, though, as Duke has mentioned. An adjustable rear-firing tweeter (plus upper midrange if necessary), would do. Sonus Faber has done something like that in the past:

View attachment 518091

Such concepts are usually resembling adjustable bipoles in the treble bands, but their sound is heavily dependent on how reflective/diffusive the rear wall behind the speakers is. I guess that was the reason why SF have implemented a mechanical possibility to alter the horizontal angle of the rear-firing unit.

Note such are concepts to widen directivity hence reducing directivity index. Adjustable directivity concepts to narrow the beam, such as the Beo 90, are much more complicated to design and usually involve driver arrays, which I would recommend for tweeters.

Very cool, didn’t know SF had such a speaker at one time. Maybe before they went the route of HEA jewelry and poor measurements.

In my case, my Soundfields use a combo of DSP, front firing planar tweeter and rear firing diffused planar to achieve the variable directivity. It works amazingly well.

 

[Bwaslo]

Similar effect to variable could be to use extra directional ambient drivers aiming to reflect off ceiling or other boundaries. Ideally with some added delay and convolved with a decent room's impulse response. (Without directivity and delay, though, it affects image and tone too much, so just a rear driver isn't a great solution).