Dutch & Dutch 8c Review

[fineMen]

You can ’tell’ the speaker how far away it is from the front and side walls via its app.
Keith

You may want to go back a bit. You'll see that we already identified the bass adjustment as being irrelevant for the cardiod operation.

 

[garbz]

We discussed earlier, that cardiods don't work that well close to a wall behind them.

Did we? Or did you just declare it so? I mean we can see quite clearly in Erin's in room measurements and his explanations of what he measured that peaks and troughs in response typically attributed to the front wall reflection are not present.
And given that basically everyone (including both D&D and Kii Audio) specifically mention this design deals precisely with the front wall reflection... You have an uphill battle with your convincing to do.

Me I'm clueless, but I can't see what you're saying reflected in the measurements here.

 

[Scholl]

Mystery is by which method the distance to the wall behind the speakers is considered in the design. We discussed earlier, that cardiods don't work that well close to a wall behind them. The caveat follows an argument directly derived from physics.

Cardioid directivity speakers work specifically better than normal speakers when we consider the wall behind them. The high-bass and low mids, which cause the most issues because of SBIR, are reduced to the rear thanks to the cardioid, while low-bass wavelength are long enough so that their reflections constructively add with the direct signal when the speakers are close to the wall. Since low bass is increased that way, we just have to slightly reduce it with a low shelf thanks to the DSP. Because bass boost will depend on the speaker to wall distance, the DSP takes that info into consideration when adjusting the low shelf.

Free bass from boundary reinforcement, less distortion, and no front wall low-mid comb filtering. D&D8C and Kii three are engineered to that very purpose.

Below is an animation from Kii Audio showing the benefit from having the cardioid directivity with regards to the front wall reflection. On the two left animations we have a normal, monopole audio source. On the right we have a cardioid one. On top the speakers are in free field, on the bottom there is a wall right behind them. On the monopole we see a strong rear wall reflection, and not with the cardioid.

The animations on the right show the wave front as it’s built up by a typical modern three-way speaker with side-firing woofers. This example is crossed over at 250Hz, time aligned and phase equalised. Indeed the wave travelling towards the listener is coherent. Towards the sides and the rear the picture breaks down and the impulse is completely smeared. Also, low frequencies get emitted in all directions. In a real room (bottom animation) the sound that escapes toward the back will strike the wall behind the speaker and reflect towards the listener, who actually hears the speaker twice: once from the front and a bit later from the rear. Moving the speaker out into the room doesn't help. The overall bass balance looks better but the impulse response gets worse as the reflection now comes even later.

 

[Kvalsvoll]

These designs are not removing all reflections, they’re removing reflections at the 5ms to 30ms window of the precedence effect. Which do nothing but muddy the sound and color it.

The decay time of good rooms is usually in the 0.25s to 0.5s range. So you still have PLENTY of reflections.

A good explanation - short, right to the point, understandable.

Though I can add that better directivity does not remove all early reflected sound, still depends on the specific design, placement, room acoustics. There will still be some mud left.

 

[fineMen]

Me I'm clueless, but I can't see what you're saying reflected in the measurements here.

Better to perform specific measurements.

Below is an animation from Kii Audio showing ...

Sure enough they take the specifics of a cardioid for granted. But as stated above, dependent on the wavelength and wall distance, that isn't a natural move.

I would rather suggest to do some calculations on Your own. And measurements as suggested - the scientific way to go. For me cardioid is the next hot thing, surely.

Regarding https://www.audiosciencereview.com/forum/index.php?goto/post&id=988091

A good explanation - short, right to the point, understandable. ...

Really, the misconception is obvious in the original posting.

 

[garbz]

Better to perform specific measurements.

Specific to what? The measurements of the directivity pattern show less energy being transmitted backwards at the frequencies where speaker distance close to the wall are most effected.
Measurements of the in room situation show frequency and time domain related problems in reference to reflection from the front wall are massively reduced.

The measurements specific or not are are in agreement that it is improving the situation, the only thing speaking against it seems to be you?

I would rather suggest to do some calculations on Your own.

At this point I'm going to call you out. You're the one making a claim here that is going against the grain, show your working. The calculations you propose other people do are quite complex. I spent many an hour sitting in front of Matlab waiting for a similar kind of simulation (e-field reflection from a metal surface) as that shown by Scholl to actually be performed, telling someone to do their own is just silly.

 

[fluid]

Mystery is by which method the distance to the wall behind the speakers is considered in the design. We discussed earlier, that cardiods don't work that well close to a wall behind them. The caveat follows an argument directly derived from physics. I'm more trained in QM, but by education I should trust my findings enough to propose a verification by measurement. The how-to was also stated.

I don't think that is a mystery, Martijn is following the work of Roy Allison and using the closeness of the wall behind the speaker to increase low bass level but still in a relatively small speaker. The distance parameter in the software would most likely change a shelving filter to keep the in room bass level as designed.

I don't agree that a cardioid doesn't work that well close to a wall, but there is nothing wrong with your acoustical mirror image descriptions.
All speakers work better close or far from walls it is the middle ground that causes trouble.

Here is an example of my W371 knock off idea as simulated in VituixCAD, it is supercardioid for most of it's intended use range. These are 15" drivers so the baffle does a lot of the heavy lifting by itself but from 300 to 100Hz the pattern is created through interference.



If you place it close to the wall behind it this is the interference response from that wall alone, the directivity is helping to make the dips quite small and boost the bass



Now if you look at the left wall where there is no attenuation from the directivity, doesn't look so good



If you add both together, most of the issues are from the left wall.



If you were to treat the left wall and let the front wall reflect the response would be pretty nice and easily correctable with basic EQ, which is I imagine the intent of the designer.

 

[fineMen]

I don't think that is a mystery, Martijn is following the work of Roy Allison and using the closeness of the wall behind the speaker to increase low bass level but still in a relatively small speaker. The distance parameter in the software would most likely change a shelving filter to keep the in room bass level as designed.

I don't agree that a cardioid doesn't work that well close to a wall, but there is nothing wrong with your acoustical mirror image descriptions.
...
Here is an example of my W371 knock off idea as simulated in VituixCAD,

Wow, I really appreciate Your effort to clarify things (and spread real knowledge)! I didn't know about that possibilities of Vituix, did it with Excell as stated before.
A) I will replicate the simulation, and look a bit further.
B) My prominent concern wasn't the dips and things, but degraded directivity to the sides; but still have to confirm if Your findings in this regard originate in the problem statement I issued

Thanks a lot, "fluid"!

 

[fluid]

A) I will replicate the simulation, and look a bit further.

This should help you
https://www.audiosciencereview.com/...rket-requirements-gathering.27212/post-938083

B) My prominent concern wasn't the dips and things, but degraded directivity to the sides; but still have to confirm if Your findings in this regard originate in the problem statement I issued

I'm not sure I understand what you mean here

Thanks a lot, "fluid"!

You are welcome

 

[fluid]

@fineMen I think I understand the point that you are trying to make regarding boundary surfaces affecting the response shape.

AKABAK has a handy pressure box feature where a specific level of cardioid rear rejection can be specified as well as the diameter of the driver which gives a good approximation of what the 8C might look like down to 100Hz.

I have made a 3D simulation of the Pressure Box in free space vs one where a fully reflective wall is placed 400mm behind it so the difference in the observation fields can be seen. There is more of change at 500 to 800 Hz, this would be dependent on the distance to the wall. Overall the pattern does not seem to be significantly deteriorated. Hopefully this is what you were getting at.

 

[Tedes Audio Pro]

I would prefer avoid flying around just for listening a loudspeaker when I can do it in Zurich.

Hello, would you consider coming to Milan to listen the 8c?

 

[bigjacko]

AKABAK has a handy pressure box feature where a specific level of cardioid rear rejection can be specified as well as the diameter of the driver which gives a good approximation of what the 8C might look like down to 100Hz.

Thanks for the simulation. How good are you at AKABAK? I am thinking is it possible to specify an arbitrary dispersion or upload our own dispersion for a driver? That would make the simulation a bit more accurate and apply to more use cases.

 

[fluid]

I am thinking is it possible to specify an arbitrary dispersion or upload our own dispersion for a driver? That would make the simulation a bit more accurate and apply to more use cases.

Yes it is possible but not necessarily simple. Here is an extract from the manual. I used the shape of the diaphragm and cardio level.

Form - Pressure Box (5000)

Page - Directivity​ "Directivity" stands for the radiation pattern of the pressure source. There are three means for providing directivity: Directivity files Lines of points Shape simulating directivity The directivity response will be weighted by all specified, i.e. accumulative. If none is specified there will be a point source radiating with constant amplitude into all directions. During calculation in a 3D-environment the pressure source has to provide directivity for any polar and any azimuthal angle. Values for in-between points are interpolated on the sphere. Directivity Files​ A Directivity File is a set of frequency response curves of sound pressure over a range of angles. Specify the reference(s) to a Contour-File-object. You can approximate a balloon response by using Directivity File(s) for the horizontal and the vertical plane. Spherical interpolation will be performed for in-between points. If you specify only one reference the balloon-interpolation will be done with Line of Points or Shape. If the latter are not specified the single Directivity File will be used for both directions. If no reference is specified the directivity defaults to Line of Points, Shape or is omni-directional. The response of the Directivity File will be normalized internally. The normalization point is the on-axis angle. Hence, after normalization the spectrum of the on-axis response is one and all the other response functions are in relation to it. The weighting of the frequency response of the Pressure Box is typically done at observation stage either with the help of the Fixed Driving Form or by means of the Lumped Element Network. Lines of points​ If Length is specified a distribution of points is installed. These points are positioned on a line or arc segment centered at the mounting point of the component. All points are driven equally by Nth part of the value. The total radiation is p = 1/N· Σi exp(-jkdi)/di with wavenumber k = ω/c and individual distance between point i and target point di. Direction​ The line or arc can lie in the horizontal or vertical plane of the component. Length​ The length of the line or the length of the secant. Arc Angle​ If specified, points would be stretched to lie on a segment of circle. Otherwise points are forming a line segment. Num Points​ Specify here the number of points on the line or on the arc. If not specified the number of points are internally calculated to be: N = Length/5mm Shape simulating Directivity​ If the size is specified this feature calculates the text-book directivity function for a flat piston in an infinite baffle. Mathematically these functions are symmetric with regard to front and rear. Asymmetry can be mimicked with the help of the combination with the Directivity File or the use of the Cardioidal Attenuation. Shape​ Shape specifies the shape of a flat piston: Disc, Rectangle, Ellipse. Diameter or Width and Height​ Size of the shape. If not specified simulation of the shape is not performed. Cardio Level Attenuation​ Simulates in a simple way the directivity for the rear radiation attempting to mimic cardioidal radiation pattern or diffraction. The value is the attenuation at the center-rear in dB.

 

[markus]

Cardioid directivity speakers work specifically better than normal speakers when we consider the wall behind them. The high-bass and low mids, which cause the most issues because of SBIR, are reduced to the rear thanks to the cardioid, while low-bass wavelength are long enough so that their reflections constructively add with the direct signal when the speakers are close to the wall. Since low bass is increased that way, we just have to slightly reduce it with a low shelf thanks to the DSP. Because bass boost will depend on the speaker to wall distance, the DSP takes that info into consideration when adjusting the low shelf.

Free bass from boundary reinforcement, less distortion, and no front wall low-mid comb filtering. D&D8C and Kii three are engineered to that very purpose.

Below is an animation from Kii Audio showing the benefit from having the cardioid directivity with regards to the front wall reflection. On the two left animations we have a normal, monopole audio source. On the right we have a cardioid one. On top the speakers are in free field, on the bottom there is a wall right behind them. On the monopole we see a strong rear wall reflection, and not with the cardioid.

Playing devil's advocat: Why not put speakers in or on a wall or even corners? Much simpler engineering, better efficiency and the speaker is out of the way (WAF).

 

[Emlin]

Playing devil's advocat: Why not put speakers in or on a wall or even corners? Much simpler engineering, better efficiency and the speaker is out of the way (WAF).

Theoretically better but in domestic settings practically impossible. Many studios use soffit mounting.

 

[markus]

Theoretically better but in domestic settings practically impossible. Many studios use soffit mounting.

How hard is it to hang a purpose-designed speaker on a wall?

 

[Emlin]

How hard is it to hang a purpose-designed speaker on a wall?

Very

 

[garbz]

How hard is it to hang a purpose-designed speaker on a wall?

Depends on if you are single or not, or if it is even your house.

 

[markus]

Very

Depends on if you are single or not, or if it is even your house.

And if it would be ON-wall? Guess something clever could be designed that still is placed on the floor. But then again people hang a lot of things on their wall, why not a speaker? Example: https://www.ikea.com/us/en/p/symfonisk-picture-frame-with-wi-fi-speaker-black-40487320/

 

[Emlin]

And if it would be ON-wall? Guess something clever could be designed that still is placed on the floor. But then again people hang a lot of things on their wall, why not a speaker? Example: https://www.ikea.com/us/en/p/symfonisk-picture-frame-with-wi-fi-speaker-black-40487320/

And where do you find the necessary volume of the cabinet?