I am happy to report that I've had very satisfying (subjective) results from a variety of music using the topology suggested by Elias. However, I could notice that any asymmetry in the nature of the walls (or lack of the same on one side) is perceived as a skew in the image. But the speaker is very tolerant of listener position, and to many of us sounded more 'enjoyable' than the rigid stereo triangle. A major difficulty posed while testing and evaluation aurally is, I feel, because of the way almost all conventional stereo recordings are made.
For quite sometime I have been evaluating 'inverted' drivers side-by-side with my simple OBs, and I had found that they were 'very good' with lots of recordings, having many of the characteristics of the OBs, for obvious reasons. So I decided to marry the two ideas and set up an L-C-R speaker, with the L & R drivers mounted 'inverted' onto two pieces of PVC pipes pushed into a shallow wooden box for bass-summing. The centre speaker was mounted onto another PVC tube with a 45 degree cut. I used the filter as suggested by Elias and fired up the system, and I must say that I was pleasantly surprised. The 'best' listening area was not up close to the unit, but at a distance, again, as observed by Elias. I intend to put in a home-made Passive Radiator (shown dotted) into the box in order to augment the bass a bit. I believe the accompanying rough sketch illustrates my setup.
@Elias: Do explain the calculation of the filter elements. Schupbach uses a 1.8 kHz 'division' in his Stereolith. What is the 'ideal' figure that we should think of ? The SSS needs to be 'tuned' to some sort of perfection in the reproduction of a 3-D sound space.
Thanks to all in the forum, and particularly to Elias, for a very "active pursuit" of an elusive quarry. SSS may not be THE answer; but it does open up a whole new possibility for the enjoyment of our hobby and what we all love-- good sound !
For quite sometime I have been evaluating 'inverted' drivers side-by-side with my simple OBs, and I had found that they were 'very good' with lots of recordings, having many of the characteristics of the OBs, for obvious reasons. So I decided to marry the two ideas and set up an L-C-R speaker, with the L & R drivers mounted 'inverted' onto two pieces of PVC pipes pushed into a shallow wooden box for bass-summing. The centre speaker was mounted onto another PVC tube with a 45 degree cut. I used the filter as suggested by Elias and fired up the system, and I must say that I was pleasantly surprised. The 'best' listening area was not up close to the unit, but at a distance, again, as observed by Elias. I intend to put in a home-made Passive Radiator (shown dotted) into the box in order to augment the bass a bit. I believe the accompanying rough sketch illustrates my setup.
@Elias: Do explain the calculation of the filter elements. Schupbach uses a 1.8 kHz 'division' in his Stereolith. What is the 'ideal' figure that we should think of ? The SSS needs to be 'tuned' to some sort of perfection in the reproduction of a 3-D sound space.
Thanks to all in the forum, and particularly to Elias, for a very "active pursuit" of an elusive quarry. SSS may not be THE answer; but it does open up a whole new possibility for the enjoyment of our hobby and what we all love-- good sound !
Attachments
I do not think the goal is the interference being "avoided" when staying
with conventional stereo triangle and modifying it by (direct) HF diffuse radiation.
http://decoy.iki.fi/dsound/ambisoni...ssment of stereophonic_Bennett et al_1985.pdf
Referring to Fig. 5 of the Bennett et al. paper i would currently describe the
"single source" and the "HF diffuse triangle" approaches as follows:
Center image
SSS: Use a real center source.
HDT: Use conventional summing localisation due to time-intensity trading.
But
Randomize interference pattern at HF in a way that off center aliases of an intended
centered phantom (e.g. critical phantom sources with high energy >1Khz) make up a "courtain"
of aliases which does not cluster around the speakers (or anywhere), as it may happen using
coherently radiating speakers at HF.
This way the interaural phase difference evaluated from low to mid frequencies gets
even more dominant (which it is usually in broadband music program) and the contradictory
information due to ITD, HRTF and interference artefacts "fades away" due to near center
images.
Side images
SSS: Make up virtual sources by sidewall reflections.
HDT: Use real side sources.
Challenges in SSS
- Virtual side sources are depending from properties of the room's side walls
- Virtual side sources are time delayed relative to the center source
- Side sources decompose into a direct and a later but more intense reflected component
Challenges in HDT
- Find a mode in going from coherent to diffuse radiation in a seamless manner(!),
which of course has to be physiologically motivated.
- Find a suitable pattern over frequency and radiation angles which builds up that
desired "courtain", even if the room does not support diffuse lateral reflections:
Sidewall reflections are assumed to help, but room requirements should be as uncritical
as possible.
Of course problems in both approaches can be mitigated by technical refinement...
Last edited:
Oliver, Again you're purely into imaging ! Where's the spaciousness ? 
Maybe it's an inherent symptom of a RFZ believer ? (what was the meaning of RFZ ??)
Yes, but as stated earlier it may not be (is not) possible without the room support: !
So, the room support is a must in order to 'randomise' the phase in stereo field. No speaker can do it alone.
Then, a simple conclusion can be made in how to generate believable phantom images:
Side Image
- SSS relies on side wall reflections
Center Image
- Stereo triangle relies on side wall reflections.
You see, fundamentally it's just looking at the same stereo problem from a different perspective
- Elias
Maybe it's an inherent symptom of a RFZ believer ? (what was the meaning of RFZ ??)Yes, but as stated earlier it may not be (is not) possible without the room support: !
So, the room support is a must in order to 'randomise' the phase in stereo field. No speaker can do it alone.
Then, a simple conclusion can be made in how to generate believable phantom images:
Side Image
- SSS relies on side wall reflections
Center Image
- Stereo triangle relies on side wall reflections.
You see, fundamentally it's just looking at the same stereo problem from a different perspective
- Elias
The room is there, so better make use of it in a beneficial way, agreed.
Nevertheless there are gradual differences in how far a system's
performance (and yes imaging) depends on room and setup.
I can agree in letting the room contribute spaciousness and also stabilize
imaging if possible, but i do not intend to hear "this or that particular
listening room".
Instead i want the room to sound better and more "understated" if you like,
than the "typical acoustical small room" would usually do.
Basically i want the room to add diffuse and sufficiently delayed cues.
That may be achieved by speaker setup and also room treatment where
necessary, but the speaker itself can bring a lot of preliminaries with it,
to come closer to that.
It is impossible to aim seriously for a concert hall in a bathroom, but you
can even excite the rectangular bathroom in a way to sound as if
it would have higher diffusivity at HF than it really has.
Like you were supposedly, i had been more in the "directional" (and coherent)
speaker camp in the past, but can mainly see advantages of that in the low to
mid frequency region now.
Stereo's deficiencies want to be "pampered" to achieve "believable" sound,
which is able to statisfy those ears who like to go to acoustical music
venues from time to time, i guess we also agree on that ...
If one visualizes first reflections at the room's walls in case of a
speaker having multiple side lobes, where the fine structure of the pattern
changes rapidly with frequency, the wall's contour will not become audible
in the way it does using a coherent speaker (say e.g. conventional omni).
You see, fundamentally it's just looking at the same stereo problem from a different perspective
- Elias
I am not able to seriously disagree ...
Last edited:
Okay, I'll be the one to ask what 95% are wondering: HDT?
(I read the cited paper. I didn't memorize or print it.)
Posters, please remember that it is not a two-person discussion, but a forum with a wide-spread audience . . . One's point is most effectively made when the audience isn't asked to be in your head.
Thanks,
-- Mark
Last edited:
Which i acronymed "HDT" (because i thought it was funny), sorry.
Meaning a conventional triangular stereo setup but with diffuse radiation
in highs somewhere above 1Khz ...
Neither single source approach nor diffuse radiation are mentioned regarding
Fig. 5 in the paper, the diagram only visualizes broadening of center images
and generation of "alias" images clustering around the speakers at higher
frequencies.
Last edited:
If one visualizes first reflections at the room's walls in case of a
speaker having multiple side lobes, where the fine structure of the pattern
changes rapidly with frequency, the wall's contour will not become audible
in the way it does using a coherent speaker (say e.g. conventional omni).
I am not able to seriously disagree ...
I think I see your point, but advocates of the omni would say that a well-placed omni is actually less revealing of wall contour, as it can (potentially) produce the most dense and uncorrelated cascade of reflections of any speaker type.
I too can't seriously agree, or disagree, without total context of room-listener-transducer details.
It's all about the total package & tradeoffs.
-- Mark
Last edited:
More comparative listening sessions have revealed something more:
My 'inverted driver' pair in the conventional stereo triangle configuration sounds very 'spacious' and images well and is tolerant of listener position, with very good localization possible even when sitting near one speaker-- probably the omni nature of the driver radiation is the reason.
Planning some more listening to the stereo triangle with a couple of added tweeters facing up and spherical reflectors on top, and then compare with the SSS.
Getting slightly confused, to admit the truth... :-(
My 'inverted driver' pair in the conventional stereo triangle configuration sounds very 'spacious' and images well and is tolerant of listener position, with very good localization possible even when sitting near one speaker-- probably the omni nature of the driver radiation is the reason.
Planning some more listening to the stereo triangle with a couple of added tweeters facing up and spherical reflectors on top, and then compare with the SSS.
Getting slightly confused, to admit the truth... :-(
Hello Prof ,
i think you have constructed an interesting "test bed" for alternative
stereo approaches, being it single source or "narrow conventional"
How far is the separation from L to R speaker ?
IMO a narrower stereo base might narrow the soundstage,
but also imaging problems due do conventional stereo triangle mitigates,
as the interference problems are shifted towards higher frequencies.
You pay width of soundstage (subjective loss of width may also depend on
the room and setup) but you get more uniform performance for listeners
sitting outside the median plane.
Even the conventional stereo triangle projects "mirror sources" of both
speakers. "Proper imaging" may not be achieved for listeners outside the
median plane, but believable "spaciousness" is still possible.
Btw. your inverted cones should behave similar like bwl, they are not "true
omni" sources. We have to expect somewhat diffuse radiation in highs due
to cone breakup (and reflection/diffraction at the basket's structure).
There is also a major portion of highs expected to be radiated towards the
ceiling. So you might have - consciously or intuitively - combined aspects
of different approaches advocated ...
Maybe you can make your system "switchable" between different modes, to
reveal the effect on different recordings.
I am curious what effect the tweeters will have.
,i think you have constructed an interesting "test bed" for alternative
stereo approaches, being it single source or "narrow conventional"
How far is the separation from L to R speaker ?
IMO a narrower stereo base might narrow the soundstage,
but also imaging problems due do conventional stereo triangle mitigates,
as the interference problems are shifted towards higher frequencies.
You pay width of soundstage (subjective loss of width may also depend on
the room and setup) but you get more uniform performance for listeners
sitting outside the median plane.
Even the conventional stereo triangle projects "mirror sources" of both
speakers. "Proper imaging" may not be achieved for listeners outside the
median plane, but believable "spaciousness" is still possible.
Btw. your inverted cones should behave similar like bwl, they are not "true
omni" sources. We have to expect somewhat diffuse radiation in highs due
to cone breakup (and reflection/diffraction at the basket's structure).
There is also a major portion of highs expected to be radiated towards the
ceiling. So you might have - consciously or intuitively - combined aspects
of different approaches advocated ...
Maybe you can make your system "switchable" between different modes, to
reveal the effect on different recordings.
I am curious what effect the tweeters will have.
Last edited:
So, the room support is a must in order to 'randomise' the phase in stereo field. No speaker can do it alone.
Then, a simple conclusion can be made in how to generate believable phantom images:
Side Image
- SSS relies on side wall reflections
Center Image
- Stereo triangle relies on side wall reflections.
You see, fundamentally it's just looking at the same stereo problem from a different perspective
- Elias
I am not able to seriously disagree ...
Although there is at least a gradual difference
:Stereo triangle can make up the center image for low to mid frequencies
quite correct (even without the room), if listener does not sit too far offset
from median plane.
Of course the problems and restrictions (due to listener position) rise with
frequency.
Last edited:
@ LineArray,
Thanks for the comments which have given me more 'food for thought' and experimentation.
For the SSS connection, the L-R separation was about 21 cm, with the L-R pvc pipes feeding the same shallow box. No, I havent yet put in the passive radiator as I thought that could wait. Actually I had been playing with the inverted drivers ever since I read about the Ohm speakers long back. But I came back to them around the time when I was fabricating my OBs. I can't prove it, but I have always had a feeling that the inverted drivers had a "different sound" than when they were conventionally baffle mounted. A major improvement was because of the lessening of the (large conventional) baffle vibrations as a result of the speaker being mounted onto a tall, narrow 'tube', whether of pvc piping or wood. Surely the masking by the magnet and frame structure and the lack of highs etc were there, but the omni sort of radiation was doing something good to the stereo imaging, I had felt.
It was easy to have a comparison between the SSS ( a la Elias) and the stereo triangle with identical drivers and mounts. For the stereo triangle, the L-R separation was more like 6 feet (my 'studio' being all of 12 feet in width, with an open slatted wall at one end, and 'broken' (not continuous) side walls on one side). With the SSS, an image skew was noticeable, though the addition of the 'steering filter' did make a difference. As soon as I was able to fire up the stereo triangle, I was 'shocked' to discover that though I was nearer to one speaker, still the spaciousness and clarity of the image was audible clearly. Replacing the jury-rigged inverted drivers with my Infinity bookshelves just changed the FR, with better highs and solid lows, but the imaging was clear only at a critical spot, and it did SOUND very much different...a bit worse, according to me! Switching back to the inverted pair gave me overall satisfaction with a variety of music (though the FR left a lot to be desired) with more 'natural sounding' voices and instruments.
The SSS, in my case, needed some lowering of the centre speaker level before the whole thing 'clicked' into place. The 'steering filter' also needs some tweaking, I guess, though I have not much idea how to go about it.
So the next logical step is to effect a low-pass well within the capability of the inverted drivers, stick a couple of 'sky-facing' tweeters onto the magnets and mount a couple of spherical ( a friend has promised a couple of stainless-steel hollow balls) reflectors on top so that there is some omni hf radiation. I have a feeling they are going to sound rather satisfactory as regards imaging and also the banishment of the sweet spot. A shorter 'pipe' and shallow box with a PR each should take care of tuning out the low-end resonances etc. It is a funny thought, but it is possible to do the whole 'plumbing' with pvc piping, a 5-6 feet long piece joining the two L-R uprights and mount one or two floor-facing passive radiators in suitable pipe couplers. Linkwitz has done it, so ... ??? (not the plumbing, but great sound with simple plumbing!)
In all honesty I must say that I have combined many approaches rather accidentally, and not consciously!!
My M-S is getting bogged down in the midst of all this ... :-(
Big thanks to all again!
-- UKP
Thanks for the comments which have given me more 'food for thought' and experimentation.
For the SSS connection, the L-R separation was about 21 cm, with the L-R pvc pipes feeding the same shallow box. No, I havent yet put in the passive radiator as I thought that could wait. Actually I had been playing with the inverted drivers ever since I read about the Ohm speakers long back. But I came back to them around the time when I was fabricating my OBs. I can't prove it, but I have always had a feeling that the inverted drivers had a "different sound" than when they were conventionally baffle mounted. A major improvement was because of the lessening of the (large conventional) baffle vibrations as a result of the speaker being mounted onto a tall, narrow 'tube', whether of pvc piping or wood. Surely the masking by the magnet and frame structure and the lack of highs etc were there, but the omni sort of radiation was doing something good to the stereo imaging, I had felt.
It was easy to have a comparison between the SSS ( a la Elias) and the stereo triangle with identical drivers and mounts. For the stereo triangle, the L-R separation was more like 6 feet (my 'studio' being all of 12 feet in width, with an open slatted wall at one end, and 'broken' (not continuous) side walls on one side). With the SSS, an image skew was noticeable, though the addition of the 'steering filter' did make a difference. As soon as I was able to fire up the stereo triangle, I was 'shocked' to discover that though I was nearer to one speaker, still the spaciousness and clarity of the image was audible clearly. Replacing the jury-rigged inverted drivers with my Infinity bookshelves just changed the FR, with better highs and solid lows, but the imaging was clear only at a critical spot, and it did SOUND very much different...a bit worse, according to me! Switching back to the inverted pair gave me overall satisfaction with a variety of music (though the FR left a lot to be desired) with more 'natural sounding' voices and instruments.
The SSS, in my case, needed some lowering of the centre speaker level before the whole thing 'clicked' into place. The 'steering filter' also needs some tweaking, I guess, though I have not much idea how to go about it.
So the next logical step is to effect a low-pass well within the capability of the inverted drivers, stick a couple of 'sky-facing' tweeters onto the magnets and mount a couple of spherical ( a friend has promised a couple of stainless-steel hollow balls) reflectors on top so that there is some omni hf radiation. I have a feeling they are going to sound rather satisfactory as regards imaging and also the banishment of the sweet spot. A shorter 'pipe' and shallow box with a PR each should take care of tuning out the low-end resonances etc. It is a funny thought, but it is possible to do the whole 'plumbing' with pvc piping, a 5-6 feet long piece joining the two L-R uprights and mount one or two floor-facing passive radiators in suitable pipe couplers. Linkwitz has done it, so ... ??? (not the plumbing, but great sound with simple plumbing!)
In all honesty I must say that I have combined many approaches rather accidentally, and not consciously!!
My M-S is getting bogged down in the midst of all this ... :-(
Big thanks to all again!
-- UKP
As far as I understand, your SSS is actually three omnis placed very close to each others ? And because of (horisontal) omni radiation pattern up to the highest freqs too, there is no 'baffle shadow' to help the high freq image steering. The idea in my SSS is at high freqs the dominant radiation is steered towards the panned side according to the stereo signal. If using just omnis, you'll lose that property.
Other than that the inverted driver omni method looks interesting, and it may well be used in an conventional stereo triangle as any other 'omni'.
Something like this (the tweeter may vary):
And if making the bottock of the driver visible, maybe could use some of the 'pretty' drivers:
- Elias
Last edited:
Naturally since the operation is based on side wall reflections !
On the other hand, I could perceive a skew in the image in an conventional stereo triangle if one of the loudspeakers is disabled
Basically stereo reproduction is very simple: You either need two side walls, or two (or more !) speakers !
It's everynone's pick which one you'll choose
- Elias
one idea: add front horn for Center driver:
- you can probably than physically time align it with L-R drivers
- having center more directive - seems to offer better imaging, while L-R will create a space by reflected sound.
Also, did some one try to direct L-R not at 90 degree to the center driver axis, but at 45 degree, to get more HF reflections from side walls and less early reflections from back wall?
- you can probably than physically time align it with L-R drivers
- having center more directive - seems to offer better imaging, while L-R will create a space by reflected sound.
Also, did some one try to direct L-R not at 90 degree to the center driver axis, but at 45 degree, to get more HF reflections from side walls and less early reflections from back wall?
@Elias: I agree with your points, but at the same time would like to point out that they are not absolute omnis, nor is the HF response that good; the centre speaker has greater level of HF and so in my case wanted some level adjustment (cut) before things were satisfactory. The omni 'mixup' did occur to me and I stuck a couple of thin board + 2" absorber of jute fibre (IMO far better than synthetic fibre wads) in the middle on both sides of the centre speaker and that improved things a lot.
One good thing I noticed with the SSS was that even with the 'skew', the image remained stable when the listener shifted position, unless of course, one moved too close to the speakers. Aurally the sound is more 'spacious and pleasing' with the inverted drivers, and the conventional stereo triangle is quite forgiving as regards listener position with the inverted drivers.
@Charcoal: My centre driver is at 45 degrees and I guess it is 'mellower' than one 'firing to your face'. In another post earlier I remember seeing the L-R speakers at about 45 degrees instead of at 90 degrees. Must try with the box with the connection suggested by Elias. As of now, my inverted drivers are taking care of more than 180 degrees on both sides, I guess, though with limited HF. Must rig up those tweeter+sphere combos soon ...
Greets to all!
One good thing I noticed with the SSS was that even with the 'skew', the image remained stable when the listener shifted position, unless of course, one moved too close to the speakers. Aurally the sound is more 'spacious and pleasing' with the inverted drivers, and the conventional stereo triangle is quite forgiving as regards listener position with the inverted drivers.
@Charcoal: My centre driver is at 45 degrees and I guess it is 'mellower' than one 'firing to your face'. In another post earlier I remember seeing the L-R speakers at about 45 degrees instead of at 90 degrees. Must try with the box with the connection suggested by Elias. As of now, my inverted drivers are taking care of more than 180 degrees on both sides, I guess, though with limited HF. Must rig up those tweeter+sphere combos soon ...
Greets to all!
Yes, but the ones I've seen were not using the SSS matrix. I'd wager that it's a sizeable step backwards. Instead of enjoying increased sidewall reflections, you'll just detect more direct L-R signal from the center (the Precedence Effect). For SSS to work as intended, the listener should only register a direct signal from the middle speaker.
Depending on driver directivity, etc, the soundstage would compress toward a "spacious mono" presentation - Nice ambiance, but narrower spread and poor panning of images.
Re: Those backwall reflections. The important thing is that the Center speaker isn't as affected by the wall behind it, unless the design calls for it (up-firing, omni, or dipole center).
If one experiences L-R trouble from the backwall, place the unit against the wall and early reflections are essentially eliminated, or address the issue just as we need to with regular 2-ch triangle: absorption on wall at the first reflection points, or increase directivity, or move the speaker farther from said wall &/or add diffusion.
An all-on-wall setup would be ideal in many respects. If one wants a more targeted sidewall reflection, simply add large flat reflective panels across the front corners angled to taste; even use the "mirror trick" to direct the L-R first reflections right to your listening position (not a good idea for larger rooms, as the delay grows too obvious without some decorrelation).
-- Mark
Here it is, the rear view:
3 cheap tactile transducers on a 3mm thick foamcore board (a dense PS foam board coated with paper on both surfaces), supported by thin wood strips, form a bending wave SSS
It's about 180 x 90 cm overall. The panel is locally damped by cork blocks and non-woven fabric pads (Thinsulate of 3M). Also, I cut 2 shallow slots to make some degree of division between 3 'zones'. The slots are not cut through, the front surface of paper coating stays intact.
The combination of these 3 transducers and foamcore have thier own problems in all kinds of horrible resonances and interferences. And it can't play low enough so additional subs are needed.
The severely flawed response is (sort of) rescued by digital EQ, then it became more listenable. (the damping also contributes)
When in use, it's not on floor as the picture, it's hung near ceiling instead. The overall presentation is very interesting. In some particular recordings, the sound stage can reach beyond its own width and go as wide as the span of original L/R chanels (more or less like the spreading of an ordinary stereo triagle).
But in most other recording, stages are narrower. The sense of spaciousness is always there, especially the depth and a sense of width at somewhere very far. The downside is the positions of images are not so well located in most cases.
The shelf filter on the central one is not used, yet. I'll try it later.
There's a function of 'stereo width' in my DEQ2496 which can enlarge the differences of the stereo 2-ch signals. I don't know what the algorithm it uses, but it does deliver the 'widening' effect. Using it moderately, it renders a big blossom picture.
Now it's not an all-rounder, yet. Far from it. It's a long way ahead. Too many things to try...
And BTW, my space is asymmetirc so I'd need additional helps from other sources (and the adjustability of them). Maybe it'd be a 5-channel system eventually...
CLS, what on earth have you done ???
Looks like fun though
I wonder, all three planes are aimed in front, and despite the 'bending' nature of sort, it propably still throws most of the energy in front too ? Maybe too narrow presentation of most stereo material. Things could improve if aiming the L and R planes more to the sides.
And, are you using matrix x = 0.5 ?
- Elias
Looks like fun though
I wonder, all three planes are aimed in front, and despite the 'bending' nature of sort, it propably still throws most of the energy in front too ? Maybe too narrow presentation of most stereo material. Things could improve if aiming the L and R planes more to the sides.
And, are you using matrix x = 0.5 ?
- Elias
Fun indeed
And yes I'm using the same matrix x=0.5 (without the shelfing filter, yet).
As to the dispersion character, this bending wave mockup (distributed mode loudspeaker) is almost omni - a bipole with very fat lobes and very narrow nulls at the sides. In ordinary rooms, it's pretty much an omni. (If this is made by 3 individual panels with either side of them turned outward, the dispersions won't be changed very much.)
So this panel is actually splashing everywhere (Not like other type of "OB" at all) When in used, there's absorbing material on the rear wall to reduce early reflection at the back.
But indeed, the 2 units on both sides are emitting too much energy to the front, and even to the 'opposite' side with equal energy. I guess it's the major drawback. Nevertheless, I still hear (feel) quite some sounds coming from the side reflections, which is clearly out of the panel itself.
This leads to a fundamental problem of my place - it's asymmetric and it shows on the sound Almost all speakers need symmetrical boundaries to perform well in sense of space, but SSS is more so.
I think I need some reflectors or diffusors to reveal the full potential of it.
And yes I'm using the same matrix x=0.5 (without the shelfing filter, yet).
As to the dispersion character, this bending wave mockup (distributed mode loudspeaker) is almost omni - a bipole with very fat lobes and very narrow nulls at the sides. In ordinary rooms, it's pretty much an omni. (If this is made by 3 individual panels with either side of them turned outward, the dispersions won't be changed very much.)
So this panel is actually splashing everywhere
(Not like other type of "OB" at all) When in used, there's absorbing material on the rear wall to reduce early reflection at the back.But indeed, the 2 units on both sides are emitting too much energy to the front, and even to the 'opposite' side with equal energy. I guess it's the major drawback. Nevertheless, I still hear (feel) quite some sounds coming from the side reflections, which is clearly out of the panel itself.
This leads to a fundamental problem of my place - it's asymmetric and it shows on the sound
Almost all speakers need symmetrical boundaries to perform well in sense of space, but SSS is more so.I think I need some reflectors or diffusors to reveal the full potential of it.