Sigberg Audio MANTA dual cardioid active speaker development thread

[Sigberg Audio]

Are there active and passive versions planned, or is this an active only project?

Active only.

 

[TLS Guy]

Sounds like an impressive system!

Okay, I will try to explain and also share measurements of what is happening, without necessarily sharing exactly how this is done.

Firstly, the ports on the front and the ports on the side have different purposes. The ports in the front baffle is as you say an attempt to "trick" the coax into seeing a more narrow baffle, to get a more even off-axis response. And it works. Here is measurements of the speaker with a closed front baffle (which sounded good already):
View attachment 55919

And here with the ports on each side of the coax:
View attachment 55920

Notice how the "lumping" at around 800-2000hz is reduced. For the record these are in-door measurements which is why they are somewhat untidy, but the point is the difference in responses.



Secondly, you have the ports on the side, which partially cancel out the sound to the sides and the rear. The most obvious way to show the effect is to compare the energy level at 180 degrees on the Manta compared to a traditional enclosure, namely our own SBS.1:

Here is on-axis vs 180 degrees for the SBS.1 (traditional enclosure):
View attachment 55921

And here is on-axis vs 180 degrees for the Manta (previous prototype iteration, the new one has yet to be measured):
View attachment 55922

Notice how the energy level is significantly reduced (actually more than 50% on average) all the way from the bottom octaves and throughout the entire audio band. The difference in how the speaker interacts with the room is almost hard to explain. Everything comes forward, you get more sense of room (depth/width) and details are easier to discern and "spread out" in 3D space. Some recordings almost sound like you're listening to a surround system. Again, hard to explain without sounding like a commercial or something I just made up.

Reduced energy towards the rear of the cabinet also reduces the cancelation of midbass frequencies due to the back wall (SBIR effects), so you get noticeably better punch in the midbass as well.

This was measured on a Klippel system with both speakers and mic fixed in place and the speaker automatically rotated, so there's no trickery of moving the speaker further away or anything involved.

Yes, you can see it works, however although the 180 degree response is significantly reduced, it is more lumpy as I would have expected. The same is true of microphones, as omnidirectional mics tend to a smoother FR, than their cardioid counterparts.

 

[TLS Guy]

Obviously early days and lots of work left, but here's preliminary on-axis + 30 degrees off-axis with the new enclosures (that arrived this week).

(0,5m in door measurement)

That is a very impressive mirror between the axis and off axis response.

 

[TLS Guy]

Active only.

And quite right too. Passive speakers are yesterday's speakers, and will be museum pieces sooner than people realize, along with their dinosaur receivers. The speakers are the correct location for power amps.

 

[Developer]

What I don't understand in this design, is how you are summing the rear cone output with the front. The output from the rear of the cone is obviously out of phase with the front.

Just so members know what we are talking about, below is the radiation patter of a driver on a flat baffle, so a di-pole.

So there are nulls at the sides. The frequency of the cancellation will be lower the greater the width of the baffle. But obviously, what you are trying to do, is to make it seem as if the baffle is narrow, and control the forward response.

However, I don't understand how your design controls the inevitable cancellations between the forward radiation and the radiation from the rear via the openings.

It seems to me that unless the rear radiation is somehow reversed to be in phase with the front radiation, then the off axis polar response will be very lumpy.

I did consider this problem around fifty years ago and solved it. It was for a system I designed for a large venue where bands and music groups performed and was for sound reinforcement.

This was an active horn system. The bass sections used back loaded 15 JBL drivers in a bass horns the size of a small kiosks, with three foot horn mouths.

The top end was a two line sources of eight Jordan Watts drivers, these were mounted in the back box. There were throats down either side, and a short exponential flare either side. This solved the reflection problem, as the width of the back box was only the width of the drivers. The horn design, inverted the rear radiation, and the flare gave just the radiation pattern I wanted.

The system sounded absolutely marvelous, and reinforced bands and vocals beautifully. Because the system was so linear there was excellent forward gain from the mics.

One evening we had a really good Bohemian band up from ND. I was in Canada then. They sounded just wonderful. I made a stereo recording of the live mix, but sent mono to the two speaker stacks, built in either side of the stage.

The evening was so good, I issued an LP of the live mix, which in the upper Midwest became a really good seller.

I think something similar might give just the response and effect you are looking for.

Changing the subject. I had not heard of those SICA drivers previously. I see they are Italian. They seem to make a huge range of drivers, with parameters from low Q high FS to higher Q but still quite high FS. Many of the specs are usually seen in PA type pro drivers. I have not had nearly enough time yet to digest that "motherload".

They claim MC Audiotech of Florida as their US agents. But MC Audiotech, make no mention of being involved in any way with SICA or their drivers.

I do not see a source of those drivers in the US, but may be you can enlighten me further on that last point, and probably the other as well.

Given that I read your post correctly, I understand that you are questioning the summation of the front and rear energy in a cardioid speaker. That is kind of a two sided question. One regards the impulse response, while the other question is about the polar pattern. The intuitive part justifies such questions, and they have been raised before. Unfortunately they are rarely approached with simple mathematics, so I will try to explain them individually.

Impulse response wise, it is important to understand the timing of a cardioid. It may be easier to cover the dispersion pattern first. One can approach it as two infinitely small sources in the space without any cabinets taking up space. The two dots can be summed easily at different angles. For a U frame, the first dot in space would be the driver, while the second dot would be the negative version of the driver with a delay that represents the distance from the back of the driver to the back opening of the cabinet. So the two are out of phase and the rear one is also delayed.

Here is a 300mm spaced version at 100Hz, followed by 200, 400, 600 and 800Hz. The two sources are equally loud:

At 300mm, half a wavelengt is 572Hz, meaning if we had no delay, the two sources would sum at +6dB (dipole peak) at this frequency. However, the rear source is delayed equally much, meaning they will be perfectly in phase at 286Hz.

We can clearly see that the wavefront is relatively intact at 400Hz, but further up it completely falls apart.

The idea of a resistive cardioid is that it partially has a U-frame shape, and partially have damping to limit the rear wave contribution at higher frequencies. With the rear source reduced by 3dB, we get this at 100Hz:

We can see that the rear damping is reduced, while the total energy is about the same. At 50Hz we can clearly see how we loose energy from the two sideds of the driver phasing each other out:

And like any dipole we get at typical dipole peak followed by a gradual drop off towards lower frequencies. This is for a theoretical cardioid without damping again:

Typically, these solutions are useful in a band of about 3 octaves, but can be extended further in some cases.

Now it becomes very simple to explain why this does not appear as two separate sources delayed in time. As you can see, the two sources are close to in phase, and the same way as a 1. order crossover, there will not be two separate waves distorting each other, they will sum as one (like summing a sine and a cosine mathematically).

The Sigberg Manta has two separate chambers. The lower chamber has the vents slightly spaced from the baffle. This increases low mid output, but makes it useless at higher frequencies. The damping is therefore focused for 100-600Hz. The top chamber has the vents on the sides, but the front vents do the most important job. The thickness of the baffle combined with the distance from the diaphragm to the front vents is actually what is needed to fuse the cardioid effect together with the natural dispersion of the upper midrange driver. Anything further back or to the sides would actually mess up the transition from cardioid to natural dispersion control from the cone.

 

[Sigberg Audio]

Yes, you can see it works, however although the 180 degree response is significantly reduced, it is more lumpy as I would have expected. The same is true of microphones, as omnidirectional mics tend to a smoother FR, than their cardioid counterparts.

Yep, but the linearity is less of an issue when the energy is reduced as much as this. Better to have it significantly reduced and a bit erratic as opposed to more linear but much higher energy.

 

[Sigberg Audio]

Here's a more graphical representation of the effect. The "traditional enclosure" graph is generated by average trend based on the spinoramas of a few similar sized speakers. Again based on the previous iteration of the prototype.

 

[Sigberg Audio]

That is a very impressive mirror between the axis and off axis response.

Here's 0-30-60-90. It's pretty flawless all the way out to 90 degrees, except something possibly going on around 1100-1400hz at 90 degrees. Remember that these are still in-doors measurements. You could be anywhere in the front of the speakers and they will sound tonally equal. All the way until you actually pass the speaker (if they're located in a way where that is possible), then you'll experience the midbass falling away due to the cardioid effect.

 

[ryanosaur]

There is precious little information I have found on Cardioid design.
@Developer or @Sigberg Audio , from a "primer-type" perspective, please:
Are there any sources you can share, please?
Is there a venting design or layout that one needs to utilize to get the Cardioid effect?
How do you determine the amount of venting area required?
What about the Resistive factor of the vents? How is that achieved?
Likewise, how is the internal compartment damped in comparison to a Sealed or Ported Cabinet (light, moderate or heavy fill? Other considerations?)?

Thank you!

 

[Sigberg Audio]

There is precious little information I have found on Cardioid design.
@Developer or @Sigberg Audio , from a "primer-type" perspective, please:
Are there any sources you can share, please?
Is there a venting design or layout that one needs to utilize to get the Cardioid effect?
How do you determine the amount of venting area required?
What about the Resistive factor of the vents? How is that achieved?
Likewise, how is the internal compartment damped in comparison to a Sealed or Ported Cabinet (light, moderate or heavy fill? Other considerations?)?

Thank you!

There's little official information on this. The original DIY thread to the guy behind the Dutch Dutch 8C is still available: https://www.diyaudio.com/community/threads/2-way-waveguide-cardioid-like.192737/

They also filed a patent:

EP3018915A1 - Directional loudspeaker - Google Patents

A directional loudspeaker is described for use in the mid frequency range of the audio spectrum. The loudspeaker comprises: a housing comprising a front panel, side panels and a back panel, said housing comprising an acoustic resistive material; wherein at least one acoustic transducer is...

patents.google.com

Neither Dutch&Dutch nor we are sharing exactly what're doing, but there's enough information out there for anyone to experiment their way there if they put in the time and effort.

There are also a few other manufacturers that are doing similar things, for instance Amphion: https://amphion.fi/products/krypton3/

 

[ski2xblack]

This approach is achieving some things I've only seen in unity or multiple entry horns. Very clever, and having heard such rigs that control response well below a room's Schrader freq., it leads to some very favorable results when it comes to interactions with local acoustics. But synergy horns are necessarily huge, at least for domestic use.

I'll also second Ryan's request for more info. References such as Toole's book focus almost exclusively on direct radiators, inexplicably give short shrift to controlled directivity methods (when some of Harman's own product, higher level LSR and Synthesis kit, would destroy the competition in those blind tests), and really nothing when it comes to these more innovative approaches such as Sigberg and D&D are doing.

Cool thread. [edit...and there's Sigberg with the links before I finished my post. Thanks!)

 

[TLS Guy]

There is precious little information I have found on Cardioid design.
@Developer or @Sigberg Audio , from a "primer-type" perspective, please:
Are there any sources you can share, please?
Is there a venting design or layout that one needs to utilize to get the Cardioid effect?
How do you determine the amount of venting area required?
What about the Resistive factor of the vents? How is that achieved?
Likewise, how is the internal compartment damped in comparison to a Sealed or Ported Cabinet (light, moderate or heavy fill? Other considerations?)?

Thank you!

After a long day, I have just got home. The information provided by this team is very interesting. The slots spacings appear to be spaced according to the delay caused by the speed of sound and also phase delays. So this would take complex math. I suspect they have generated a computer design model.

Whether there is damping in the back box, they did not state. If there is, it must be minimal, or the rear radiation would be adsorbed, and none would be emitted from the ports, to create the required interference for the cardioid effect. I'm still a bit of a sceptic about the front port radiation though.

One thing that does concern me is reflections from underdamping, or none, causing reflections through the cone. The cone would have to be well damped inherently to limit this. This might well preclude the use of metal cone drivers, as you have to make sure you damp the rear radiations behind the driver, as these are freely transmitted through metal cones, because of the high Young's modulus.

I find this approach really interesting, and will follow closely. I really appreciate this team educating us in the way they do. This reminds me of the pioneers of years ago, in the fifties and sixties, who would never keep advances secret and always eager to educate. Keeping anything to themselves would have been totally contrary to the ethics, of the likes of Peter Walker, Jim Rogers, Stan Kelly, and Raymond Cooke, and many others. Secrets create suspicion. Sharing new ideas and proving your point generates sales and enthusiasm. People of the ilk I mention above just hated marketers and their trite, confusing bilge.

So, I hope these new endeavors push the needle forward, and full disclosure generate sales, which they should.

 

[TLS Guy]

There's little official information on this. The original DIY thread to the guy behind the Dutch Dutch 8C is still available: https://www.diyaudio.com/community/threads/2-way-waveguide-cardioid-like.192737/

They also filed a patent:

EP3018915A1 - Directional loudspeaker - Google Patents

A directional loudspeaker is described for use in the mid frequency range of the audio spectrum. The loudspeaker comprises: a housing comprising a front panel, side panels and a back panel, said housing comprising an acoustic resistive material; wherein at least one acoustic transducer is...

patents.google.com

Neither Dutch&Dutch nor we are sharing exactly what're doing, but there's enough information out there for anyone to experiment their way there if they put in the time and effort.

There are also a few other manufacturers that are doing similar things, for instance Amphion: https://amphion.fi/products/krypton3/

I really appreciate all this detailed information. I have thought about these problems. That is why my main speakers are the shape they are with the swept back fronts, and the design of the room, specifically the design of the front of the theater and the way the let and right speakers are placed. Obviously an approach like that has to be custom. However the FR is linear on and off axis, and the room curves excellent. That design system however is not suitable for the market, but a definite perk of the DIY approach.

 

[Sigberg Audio]

Whether there is damping in the back box, they did not state. If there is, it must be minimal, or the rear radiation would be adsorbed, and none would be emitted from the ports, to create the required interference for the cardioid effect. I'm still a bit of a sceptic about the front port radiation though.

One thing that does concern me is reflections from underdamping, or none, causing reflections through the cone. The cone would have to be well damped inherently to limit this. This might well preclude the use of metal cone drivers, as you have to make sure you damp the rear radiations behind the driver, as these are freely transmitted through metal cones, because of the high Young's modulus.

Both chambers are actually more than adequately damped to avoid cabinet reflections. All surfaces are covered by a composite felt material similar to the SBS.1 (if you saw that review), and there is also additional dampening going on.

The picture below gives an indication (this is not the final dampening situation, but shows the surfaces being mostly covered:

 

[TLS Guy]

Both chambers are actually more than adequately damped to avoid cabinet reflections. All surfaces are covered by a composite felt material similar to the SBS.1 (if you saw that review), and there is also additional dampening going on.

The picture below gives an indication (this is not the final dampening situation, but shows the surfaces being mostly covered:

View attachment 55946

Once again, thanks for those pictures. I suspected it was probably damped along those lines, rather then stuffed with damping material as usual practice in not tuned boxes.

This is an interesting design, and I will follow closely. Please keep the info coming. The time for novel active speakers is now. Passive speakers are at the end of their run, they will soon be history. So Yamaha and Sound United need to be gearing up to produce affordable AV preamps. They won't need to waste money on Audyssey, as the correct an best place to do this will be in the speaker. We just don't need huge boxes with eleven or thirteen power amps!

 

[shadyJ]

Passive speakers are at the end of their run, they will soon be history. So Yamaha and Sound United need to be gearing up to produce affordable AV preamps. They won't need to waste money on Audyssey, as the correct an best place to do this will be in the speaker. We just don't need huge boxes with eleven or thirteen power amps!

I do think that there will always be a place for passive speakers, although active speakers are becoming quite popular. Passive speakers do not have electronics onboard, so they will likely last a lot longer. Also, active speakers need a power connection for every single speaker, and that is more power cords for your outlets. sometimes that is not feasible. Obviously, active speakers have their advantages, but so do passive.

 

[Developer]

After a long day, I have just got home. The information provided by this team is very interesting. The slots spacings appear to be spaced according to the delay caused by the speed of sound and also phase delays. So this would take complex math. I suspect they have generated a computer design model.

Whether there is damping in the back box, they did not state. If there is, it must be minimal, or the rear radiation would be adsorbed, and none would be emitted from the ports, to create the required interference for the cardioid effect. I'm still a bit of a sceptic about the front port radiation though.

One thing that does concern me is reflections from underdamping, or none, causing reflections through the cone. The cone would have to be well damped inherently to limit this. This might well preclude the use of metal cone drivers, as you have to make sure you damp the rear radiations behind the driver, as these are freely transmitted through metal cones, because of the high Young's modulus.

I find this approach really interesting, and will follow closely. I really appreciate this team educating us in the way they do. This reminds me of the pioneers of years ago, in the fifties and sixties, who would never keep advances secret and always eager to educate. Keeping anything to themselves would have been totally contrary to the ethics, of the likes of Peter Walker, Jim Rogers, Stan Kelly, and Raymond Cooke, and many others. Secrets create suspicion. Sharing new ideas and proving your point generates sales and enthusiasm. People of the ilk I mention above just hated marketers and their trite, confusing bilge.

So, I hope these new endeavors push the needle forward, and full disclosure generate sales, which they should.

I just want to point out that I am not directly involved in this project, but I have provided a few simulations early in the project, just to provide the link between the maths and the positioning of the driver and slots. I also recommended the Keyser-thread on DIYAudio for reference. I also know Mr Sigberg, and we often discuss measurements and technology. And I have heard the first Manta protos. That was quite an experience.

I think you got the part with metal cones and damping wrong. Sound from inside the cabinet will typically penetrate through the diaphragm quite easily, almost regardless of the type of diaphragm. If you want proper midrange, you need the diaphragms to have low loss. Such diaphragms can not stop sound effectively from passing straight through. With extremely high damping, which can work for subwoofers, you can stop sound quite effectively, but that will sound extremely bad used as midrange.

 

[Sigberg Audio]

In other news, the new prototype has got revised side ports, especially the lower chamber has significant adjustments. Measurements have only been done indoors so far, so this won't tell us energy loss in absolute numbers due to reflections. What we can do is see the difference between the previous and the new prototype. Something is clearly going on from 150hz and below that looks very interesting.

Green is on-axis, red is 180 degrees on the old prototype, and blue is 180 degrees on the new / current prototype.

 

[TLS Guy]

In other news, the new prototype has got revised side ports, especially the lower chamber has significant adjustments. Measurements have only been done indoors so far, so this won't tell us energy loss in absolute numbers due to reflections. What we can do is see the difference between the previous and the new prototype. Something is clearly going on from 150hz and below that looks very interesting.

Green is on-axis, red is 180 degrees on the old prototype, and blue is 180 degrees on the new / current prototype.

It suggests that there is in phase augmentation
Between the front radiation and the ports at 225 Hz, and excessive cancellation by out of phase at 150 Hz.
This has the hall marks of a project that could drive a guy insane!

 

[Sigberg Audio]

It suggests that there is in phase augmentation
Between the front radiation and the ports at 225 Hz, and excessive cancellation by out of phase at 150 Hz.
This has the hall marks of a project that could drive a guy insane!

As mentioned this is in-room measurements, I suspect what we're seeing at 225hz is not a real problem. It's present in the old prototype (red measurement) as well, and this was not seen during the Klippel measurement of that speaker. The point of the graph was the fact that the energy has been reduced significantly below 150hz.