In room measurements of Beolab 50?

[Bjorn]

Another consideration is that "cardioid' speakers usually approach the ideal cardioid pattern only within select bands.

Very true and is also what I have shared in this thread and said it's fairly constant. Often, the directivity is more in the area of 140-160°. So about 10-20° less to each side compared to a traditonal 180° dispersion speaker. That's something but not a lot.

The real advantage, in my opinion, in that a well designed controlled directivity loudspeaker does so down to the modal region of the room in which it is being used. And to me this is audible when the first lateral reflection matches the direct sound as close as possible within a certain passband. I'm not sure it is measurable, in my room at least I don't believe it is, though I'd have to verify. My subjective feeling is that there is a more 3 dimensional aspect (depth) to the presentation when this is the case, but you can blame my imagination of course.

You can achieve the same thing with a well designed wide baffle of course.

Sure, a more uniform response result in that. A wide baffle can be more uniform and depended on the design it will narrow some in the highs.

The R2 project for instance, the goal for me was always a non-collapsing polar, not cardioid response in itself. Once we achieve that, then we simply calculate where the floor bounce will occur, and solve that through means of a seperate bass module. The only thing not addressed is further vertical directivity control.

The vertical is a weakness of most cardioid speakers and one reason they don't measure that well in-room compared to some other designs.

If I was going to design a cardioid speaker, and yes I have considered it, I would prefer using 15" woofer and active cardioid. While the combination of a speaker with smaller footprint and a fairly constant directivity goes out the window, there are clear benefits. And my interest is more in high-end audio. I think today's popular cardioid speakers sound ok and generally better than passive speakers with a poor directivity and often somewhat uneven on-axis response too. But they don't present high-end audio for me.
If you, however, combine a 15" woofer in active cardioid and can do something about the vetical, than we are more in the high-end territory with trully great sound quality.

 

[Bjorn]

I recall @napilopez posting in-room measurements of the D&D 8C a few years back compared to the Kef R3 and B&W something, and the frequency response was much better. Of course, it's hard to isolate the effect of cardioid dispersion specifically against the effect of all the other differences...

In a study that was conducted many years ago in a small room between a monopole, dipole and cardioid the overall result is the same with optimal placement for each one. However, a cardioid was considered slightly less dependent on placement which makes sense above a certain frequency area. I have to see if I can find that study again when I get time.

So if one is stuck with a certain placment, one can be better than the other. A dipole for example, will ofte thrive far our from the front wall and close to the side walls. And the dipole cancellation will trigger less room mode simply because it doesn't go as low.

 

[sigbergaudio]

One of the main perceived differences that I find is a larger and more holographic soundstage, and higher overall clarity. And this is something we don't have a clear understanding of how we measure (or analyze from measurements). Which makes it difficult to "prove" from in-room measurements to which degree it is beneficial.

 

[sigbergaudio]

Here are SBS (traditional sealed) and Manta (cardioid), single speaker measurements. Speakers in the same position measured from the same listening position.

The differences are not always as clear as this, it depends on the room and the placement. But the difference in soundstage and perceived size are present irregardless if one can clearly see measured improvements in the 100-500hz area or not. And of course the Manta has a wider baffle and larger driver, so the differences can have a combination of causes.

 

[thewas]

It's not a wide dip, it's quite narrow. Overall the response is very similar.

I don't see it this way, anyway repeating our opinions won't bring anything, let's leave each reader decide for himself.

While you can EQ peaks, the result is withour compromise when it's not minimum phase and the time domain isn't improved and actually made worse by phase distortion.

I agree, but peaks in that region are usually mainly minimum phase.

 

[thewas]

Here are SBS (traditional sealed) and Manta (cardioid), single speaker measurements. Speakers in the same position measured from the same listening position.

The differences are not always as clear as this, it depends on the room and the placement. But the difference in soundstage and perceived size are present irregardless if one can clearly see measured improvements in the 100-500hz area or not. And of course the Manta has a wider baffle and larger driver, so the differences can have a combination of causes.

View attachment 483745

Thank you, so the non-cardioid shows a similar wide SBIR dip in the midbass like on my measurements while the cardiod doesn't.

 

[Snickers-is]

If I was going to design a cardioid speaker, and yes I have considered it, I would prefer using 15" woofer and active cardioid. While the combination of a speaker with smaller footprint and a fairly constant directivity goes out the window, there are clear benefits. And my interest is more in high-end audio. I think today's popular cardioid speakers sound ok and generally better than passive speakers with a poor directivity and often somewhat uneven on-axis response too. But they don't present high-end audio for me.

This sounds very strange to me.

Firstly, a purely horizontal cardioid will also work vertically. It is just simple geometry. The pattern is not identical, but quite close.

Secondly, there are several studios that has reviewed, and ended up purchasing Sigberg Audio's cardioid speakers. I have even been present myself in a couple of those cases, hearing the result, not to mention hearing the technicians response. I mean, come on, what they say is that the cardioid speakers wipe the floor with their competitors. But what do they know, they are just simple workers, right?

You also talk about some huge advantages with non cardioid speakers. But you do not mention what those might be. This sounds pretty strange to me, what are you aiming at here?

 

[sigbergaudio]

Full range stereo (left+right average) of the Manta (+ subs of course) in the same room. Depending on placement the SBIR in the 100-300hz area can be avoided more or less completely.

 

[kharan]

The difference in radiation of cardioid speakers is well documented. So what remains is what actual perceived differences this translates to for the listeners in any given room. That is a pretty complicated question.

In my designs I try to amplify the effect by also designing the speakers to be placed close to the wall, which further reduce the SBIR effects.

I also think cardioid is beneficial in higher frequencies, and here we typically actually see smoother in-room response in the higher frequencies with the Manta than for instance the SBS, which is a bit interesting since the SBS is more tidy anechoically.

Can you define higher frequencies? From all the measurements I've seen, above maybe 700 Hz, it's usually not too onerous both from a space and aesthetic standpoint to design speakers with a relatively constant directivity with either both narrow or wide dispersion in the horizontal axis. Companies have achieved this with "merely" a sensible choice of driver, crossover and enclosure design.

Of course, this is only when said speakers are designed to achieve controlled directivity over that entire range, which is the case for the aforementioned Neumann.

That leaves us with the upper bass lower mids range, which I think cardioid radiation can help control without deviating too far from traditional speakers looks.

EDIT: I hadn't seen your SBS post about the 100-500 Hz range yet.

 

[sigbergaudio]

Can you define higher frequencies? From all the measurements I've seen, above maybe 700 Hz, it's usually not too onerous both from a space and aesthetic standpoint to design speakers with a relatively constant directivity with either both narrow or wide dispersion in the horizontal axis. Companies have achieved this with "merely" a sensible choice of driver, crossover and enclosure design.

Of course, this is only when said speakers are designed to achieve controlled directivity over that entire range, which is the case for the aforementioned Neumann.

That leaves us with the upper bass lower mids range, which I think cardioid radiation can help control without deviating too far from traditional speakers looks.

EDIT: I hadn't seen your SBS post about the 100-500 Hz range yet.

Well, the Manta has two cardioid systems, and the upper one starts at 600hz.

If we look at 0 vs 180 degrees of the Manta vs the SBS (non-cardioid), the attentuation at 180 degrees is quite a bit higher up to around 5khz. Parts of this is due to the larger speaker of course. But early tests, where otherwise identical cabinets with and without ports were compared, significant differences in radiation up to between 1500 and 3000hz were measured (depending on the angle).

Note that the graphs below are from prototypes, so this is not identical to the production versions, but close enough for comparison purposes.

Manta:

SBS:

 

[kharan]

Well, the Manta has two cardioid systems, and the upper one starts at 600hz.

If we look at 0 vs 180 degrees of the Manta vs the SBS (non-cardioid), the attentuation at 180 degrees is quite a bit higher up to around 5khz. Parts of this is due to the larger speaker of course. But early tests, where otherwise identical cabinets with and without ports were compared, significant differences in radiation up to between 1500 and 3000hz were measured (depending on the angle).

Note that the graphs below are from prototypes, so this is not identical to the production versions, but close enough for comparison purposes.

Manta:
View attachment 483754

SBS:
View attachment 483755

That's more due to how the SBS is designed imho. You have achieved high attenuation, around 25 dB from 1500 to 3000 Hz at 180 degree with your cardioid design in the Manta. But other designs can achieve very similar -25 dB or even -30 dB in the same range and angle.

KH 420 with a waveguided medium and tweeter.
Sourcepoint 888 with driver and enclosure design
S360 with a horn
JBL Studio 698 also with a horn, but the design is a very classical looking three way.
Grimm LS1 is also almost there with with its wide design, and they don't use any kind of deep waveguide or horn.

All of those speakers (data available on spinorama.org) present similar attenuation at 180° to the Manta, all with very different looks and construction.

 

[sigbergaudio]

That's more due to how the SBS is designed imho. You have achieved high attenuation, around 25 dB from 1500 to 3000 Hz at 180 degree with your cardioid design in the Manta. But other designs can achieve very similar -25 dB or even -30 dB in the same range and angle.

KH 420 with a waveguided medium and tweeter.
Sourcepoint 888 with driver and enclosure design
S360 with a horn
JBL Studio 698 also with a horn, but the design is a very classical looking three way.
Grimm LS1 is also almost there with with its wide design, and they don't use any kind of deep waveguide or horn.

All of those speakers (data available on spinorama.org) present similar attenuation at 180° to the Manta, all with very different looks and construction.

Yes, and again the main benefit beyond reduced SBIR in the lower midrange is a larger and better perceived soundstage and clarity, which we don't really know how to reliably measure or identify in measurements.

And yes there are certainly different ways to achieve rearwards attentuation.

 

[kharan]

Yes, and again the main benefit beyond reduced SBIR in the lower midrange is a larger and better perceived soundstage and clarity, which we don't really know how to reliably measure or identify in measurements.

And yes there are certainly different ways to achieve rearwards attentuation.

Just flush mount a pair of 8380, no more rearwards radiation, no more baffle step, and no more edge diffraction either.

There might ancillary effects though.

 

[MAB]

Would be interesting to see your measurements as a starter.

Agreed.
I am even more interested in seeing measurements the VA Coherence 12.
A complete set of measurements.
Like from Erin or Amir.
Since this seems like this thread is more than just a Beolab speaker design question.

Maybe then we could compare to other models and design concepts.

 

[Snickers-is]

There are some measurements of those on the forum, but they are extremely smoothed with like 0-120dB of Y scale, so it is hard to compare them. I believe it was also done on a more reasonable distance so the drivers has some time to blend. That is probably very critical since the mid and the tweeter are so far apart, even for a relatively low crossover point. Not sure what the distance used is, since that far spacing wold mean a crossover reference phase axis starting about mid way between the drivers would tilt upwards. The tweeter is very tall so one wold need to pick the point where the filter reference phase axis crosses the tweeter's vertical null axis. The B&O is probably at least a three way, so it should also be measured at a longer distance to make it fair to the product.

 

[kharan]

I understand the Klippel NFS can deliver anechoic at 2m if that's what you ask it to do.

 

[Juhazi]

Below Schroeder (3-400Hz) we see room modes and first reflection effects - peaks and dips and possibly pressurization. Different radiation pattern types vary in most speakers, only dipole panels are (nearly) constant, but need heavy EQ. I don't see point to rate patterns, each has pros and cons. Floor reflection null looks bad but brains adapt to that - straight measured response sounds boosted!

Subwoofer range below 100Hz is mostly about room modes, 100-300 about first reflections and weak secondary modes. Speaker and listener/mic placement is very delicate, acoustic panels and absorbers/traps work well if they can be used. IMO signal EQ should be done for lowest two room modes and general tonality only.

Some normalized horizontals from Erin's Klippel NFS



My diy AINOgradient is omni below 100Hz, cardioid 100-250 and dipole above that. Here outdoor normalized horizontals and room MMM after mild room EQ, with speakers, sofas and table etc. at normal locations. Notice L/R difference from lateral room modes! I can get almost same room curves with normal tower speakers and EQ but they don't sound similar in stereo and two ears...

 

[Juhazi]

About different speakers sounding different, we should also look at room decay. REW has many different types and adjustable parameters.

Here some wavelet samples in my living room, single speaker measurements at same speaker and mic location. ER18DXT is two-way BR on a stand, MR183w is 3-way tower with downfire woofer

 

[Bjorn]

Agreed.
I am even more interested in seeing measurements the VA Coherence 12.
A complete set of measurements.
Like from Erin or Amir.
Since this seems like this thread is more than just a Beolab speaker design question.

Maybe then we could compare to other models and design concepts.

While polar measurements will eventually be shared of the VA Coherence 12, I am of the opinion that most will benefit more from looking at in-room measurements.

This is were one will see that the speaker measures generally more even above 100-120 Hz compared to most speakers on the market.

 

[ssashton]

Perhaps you could do simulations in multiple room environments, to show it? Easier than a physical setup in many different rooms.