B&O Beolab 90 - adjustable directivity by DSP

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
So B&O just released their funky new flagship. Uses DSP to shape dispersion and yields excellent polar performance.

BeoLab 90: Behind the scenes earfluff and eyecandy

B&O Tech: What is “Beam Width Control”? earfluff and eyecandy

Beamwidth can be switched with a touch on their app, with the ability to configure beamwidth presets that one can switch in between on the fly. Comes with automated room correction and 18 channels of amplification/channel too. Can anyone explain how the unusual driver array works? I'm guessing that some drivers are deliberately aligned out-of-phase digitally to induce cancellation and shape the dispersion pattern that is desired.
 
similar to the kii. Look here for an animation on how they do it acoustics

I've actually read up on the Kii. Similar tech, but different priorities. Putzey's aim is to use the cancellation to significantly reduce backward radiation below baffle step, giving an effectively cardiod dispersion pattern in the bass, except for the lowest frequencies. The midrange and treble is relatively textbook (for an active) to my understanding. And the dispersion pattern is fixed.

B&O's design is clearly more ambitious. It aims for near-CD performance with variable beamwidth from direct radiators. What I struggle to understand is how the directivity can be varied so much between different DSP profiles?
 
Member
Joined 2014
Paid Member
It's a phased array. By driving each unit from a separate amplifier you can vary the phase between them. As its DSP you can then do tricks to change relative phasing against frequency. Not entirely dissimilar to how the Quad ESL 63 replicates a point source from a big flat sheet of cling film.

Whether you can make a silk purse from a sow's ear is of course another question :)
 
I fail to understand how the polars can look as constant as it does on page 15 and 16 (narrow and wide) when multiple drivers are used in a cluster. That would normally lead to lobing in both planes. Would be nice to see independent measurements.
 

Attachments

  • bl90_reflection_01.jpg
    bl90_reflection_01.jpg
    52.1 KB · Views: 1,171
  • bl90_prototype4a_01.jpg
    bl90_prototype4a_01.jpg
    64.4 KB · Views: 1,169
I fail to understand how the polars can look as constant as it does on page 15 and 16 (narrow and wide) when multiple drivers are used in a cluster. That would normally lead to lobing in both planes. Would be nice to see independent measurements.

Yes! That's what I'm perplexed about. Horizontal directivity, I can understand due to perhaps the properties of the phased array (judicious cancellation and delay to yield smooth directivity characteristics) and B&Os DSP. Vertical directivity? I can only imagine large amounts of lobing.

EDIT: Geoff Martin, the designer, provides further details of the DSP setup:

'The point of the rear-facing drivers is to interfere with the front-firing drivers. However, this is done in a very controlled fashion using an individual FIR filter customised for each driver. Consequently, at one frequency, the back drivers may be cancelling the front in a given direction whereas at another frequency, they are “helping”. The magnitude and phase responses of the filters are the result of the measurements of the drivers in their locations, and an optimisation algorithm designed to find the best possible solution to the target of a given beam width.'

The unique selling point as I see it is how automated the entire process is.
 
Last edited:
Horizontal directivity, I can understand due to perhaps the properties of the phased array (judicious cancellation and delay to yield smooth directivity characteristics) and B&Os DSP.
Can you explain further how this is accomplished?

I assume the crossover between the tweeters and mids is fairly high. Meaning the spacing between them in this freq. area will cause lobing. Unless you use a very large amounts of drivers (like in a line-array) you will have lobing and comb filtering.

Omni is my opinion a waste to use indoor in a living room. It's more than sufficient to have a wide dispersion which covers all the listeners no matter where they sit. Making it omni only causes unwanted frontwall reflections.
 
Can you explain further how this is accomplished?

I assume the crossover between the tweeters and mids is fairly high. Meaning the spacing between them in this freq. area will cause lobing. Unless you use a very large amounts of drivers (like in a line-array) you will have lobing and comb filtering.

Omni is my opinion a waste to use indoor in a living room. It's more than sufficient to have a wide dispersion which covers all the listeners no matter where they sit. Making it omni only causes unwanted frontwall reflections.

I meant to say that smooth horizontal directivity is IMO fairly achievable with such a design, but I doubt the vertical performance. To put it another way, vertical lobing is a bigger problem with this design in my opinion.
 
I know you referred to the horizontal, but I don't see how it's much different than the vertical. As you can see from the picture below, the drivers are also placed next to each other. They are placed in a triangular fashion that minimizes the spacing a little but it doesn't solve the problem.
attachment.php



So normally, even with the narrow setting, there would clearly be horizontal lobing and combing. This will increase when choosing "wide" and using the drivers on the side. But both the polars look great. Either they have solved this somehow or smoothing is covering it.
 

Attachments

  • bl90_reflection_01.jpg
    bl90_reflection_01.jpg
    52.1 KB · Views: 1,736
Bjørn:
First off all, I think you are right about the vertical response. It has to be somehow affected by the number of drivers. However, the energy response in such a setup will benefit from the kind of dispersion control you have in the horizontal plane.

So if we look at the horizontal dispersion control to begin with, you can say that lobing will occur if drivers are spaced far apart, but then again, you only need 3 drivers to get perfect dispersion control at one frequency. So what BO is doing here is that they have two narrow band dispersion control speakers facing forward, one with midranges and one with tweeters. The three forward facing tweeters will work together to control the dispersion in a relatively narrow band, much like the first 3 rings of the ESL63. The band that is not covered by this is at a significantly lower frequency band. This means you can successfully extend the dispersion control with the additional tweeters on the side without running into serious horizontal loobing problems. The midrange array is slightly larger and can therefore perform the same task at lower frequencies.

Then if you imagine a mic in front of the speaker, and you lift it upwards you will off course get lobing. If you place the same mic behind the speaker, you will get almost nothing at all. If you start lifting it, you will need to lift it pretty high up before it starts picking up a significant amount of energy from the front. This is a good indicator that the energy response is pretty nice too. However, how nice it actually is will be up to the interaction between the vertically spaced tweeters and midranges.
 
Legacy do a bit of directivity control too. I think the B&O is the first to use separate amping of every driver and advanced DSP to get it all integrated right.

One thing though - they say each speaker has 8KW of amps. But you can only draw a max of 3KW from a mains socket ;)

The Kii Three does it too: Kii Audio GmbH - High End Active Speakers for HiFi and Professional use.

All that power is there only for headroom. 8 kW amps doesn't mean it will draw that much power from the mains socket. Only if all amps were to deliver maximum output over a sustained period of time are you likely to run into trouble. You'll probably fry one or more of the drivers before exceeding the maximum power that can be supplied to the amps.
 
Does anyone here know what exactly the differences are between the B&O concept and the Ken Kantor AR Magic Loudspeaker (yes, 1985!).

BTW: what I found rather ironic is that the starting point of the current design, is a most conventional WWMTWW speaker, a design violating all controlled directivity principles..

Regards,

Eelco
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.