I have searched many cbt designs but so far all cbts suffer from poor (uneven) horizontal directivity. The baffle makes that inevitably, and I would gess that even L and R are different in most designs!
Selah audio Symmetrica is the best I have seen. Keeles original CBT is much worse.
a cbt's virtue is in vertical uniformity and I believe when many people say they sound fine. But if we look at room (power) responses in actual rooms, many systmes can give a good result. The difference is in distribution of sound energy in 3-D and reflections' delays -> imaging. It is wise to suppress directions that make early reflections.
Loudspeakers have been R&D'd for a century now, yet we don't have an overall winner.
Practically, room and loudspeaker(s) make a system and we have (way too) many details that must be considered! I don't expect any solution to be perfect. This is the thing that makes this hobby an endless endeavour
I'm glad that I'm not a pro!
Selah audio Symmetrica is the best I have seen. Keeles original CBT is much worse.
a cbt's virtue is in vertical uniformity and I believe when many people say they sound fine. But if we look at room (power) responses in actual rooms, many systmes can give a good result. The difference is in distribution of sound energy in 3-D and reflections' delays -> imaging. It is wise to suppress directions that make early reflections.
Loudspeakers have been R&D'd for a century now, yet we don't have an overall winner.
Practically, room and loudspeaker(s) make a system and we have (way too) many details that must be considered! I don't expect any solution to be perfect. This is the thing that makes this hobby an endless endeavour
I'm glad that I'm not a pro!
Who does? 🙄 Actually I think that a synergy horn is in theory the best solution. The sad thing is their size when thinking of domestic listening rooms.
Last edited:
I don't know if other have had the same experience, but I find that the directivity of tractrix horns is well controlled (almost constant, 60° beamwidth) in a small range comprise between 5 times and 15 times the cutoff of the horn (not even two octaves...). So for a 400Hz tractrix horn that would be is 2kHz to 6kHz...
Of course it can be used under 2kHz (and it will probably be easier to cross with a 12" woofer around 1kHz than a real CD device...), but it requires a tweeter above 6kHz if not directly on-axis...
A 1.2kHz tractrix horn would be a good complement (6kHz-18kHz constant range!)
Regarding dynamic sensation of a horn vs a waveguide, or more generally a shallow horn vs a deeper one, I have had the same feeling: with two devices side by side (shallow JBL PT waveguide vs somewhat deeper JBL H9800 horn) it "looked" like the deeper horn gave a better sense of dynamic, but I fail to understand or to measure why...
That is not due to distortion (way down in level for both units), directivity pattern (nearfiled listening), or tonal balance (both units where precisely EQed for the same on-axis response). Both units where also used in the frequency range that suited the smallest unit, and of course the same compression driver type (from a matched pair) were used.
😕
I think that depends on the tradeoffs that you want. Waveguides will have a smoother response than a synergy horn, but will have a single vertical lobe at the crossover. Synergy horns would also array well, which is not an issue in homes.
Did you look at the polar I posted? That's an independent measurement done under right conditions.I have searched many cbt designs but so far all cbts suffer from poor (uneven) horizontal directivity. The baffle makes that inevitably, and I would gess that even L and R are different in most designs!
Selah audio Symmetrica is the best I have seen. Keeles original CBT is much worse.
a cbt's virtue is in vertical uniformity and I believe when many people say they sound fine. But if we look at room (power) responses in actual rooms, many systmes can give a good result. The difference is in distribution of sound energy in 3-D and reflections' delays -> imaging. It is wise to suppress directions that make early reflections.
Loudspeakers have been R&D'd for a century now, yet we don't have an overall winner.
Practically, room and loudspeaker(s) make a system and we have (way too) many details that must be considered! I don't expect any solution to be perfect. This is the thing that makes this hobby an endless endeavour
The CBT36 vs. B&W is conducted in a regular room with no treatment, uenven floor and with 1/12 smoothing. Don Keele may have done a mistake publishing measurements like that. People compare it to anechoic or nearfield measurements with much more smoothing.
From what I understand the CBT36 will actually measure a little better horizontally then the prototype and the CD will go much lower.
Bjorn
I found the polars hard to read because they were so small. What is the scale? Things do look a little rough above about 5 kHz.
I agree with you that publishing real data is dangerous because it becomes target. You never really get much traction from them. But not doing so just puts us back into the Dark Ages of audio where magazine subjective reviews were the standard since no one ever looked at any real data.
It is also pointless to try and compare systems done with different presentation schemes and I believe that the polar map is the best way (as does Genelec and others.)
At any rate, even if the CBT does not collapse in the horizontal at HFs (but then neither does a waveguide) it is still way too wide IMO. I would place the CBT on its side if I had a pair. I just don't buy that lots of early reflections are a good thing.
I found the polars hard to read because they were so small. What is the scale? Things do look a little rough above about 5 kHz.
I agree with you that publishing real data is dangerous because it becomes target. You never really get much traction from them. But not doing so just puts us back into the Dark Ages of audio where magazine subjective reviews were the standard since no one ever looked at any real data.
It is also pointless to try and compare systems done with different presentation schemes and I believe that the polar map is the best way (as does Genelec and others.)
At any rate, even if the CBT does not collapse in the horizontal at HFs (but then neither does a waveguide) it is still way too wide IMO. I would place the CBT on its side if I had a pair. I just don't buy that lots of early reflections are a good thing.
This goes right to the heart of what this thread is all about.
Some combinations of flare profile, mouth area/length, and front/rear chambers provide better acoustic loading than others. This is directly seen in the acoustic impedance, and indirectly seen in the electrical impedance and amplitude and impulse response.
Strong acoustic loading provides more resistance and less reactance, and the end result is smoother response. Devices that are more reactive are more resonant, and so they are only providing a strong acoustic load at certain peak frequencies. In between those frequencies, the acoustic load is weak and this vascillation creates ripples in impedance and response.
Likewise, abrupt transitions inside a horn cause discontinuities that are also seen in impedance and response. To me, these look like similar wavefront distortions, shifted higher in frequency because they are features of a smaller portion of the horn. For example, we see the mouth termination discontinuity causing ripple near lower cutoff, any secondary flare or radius for waistband mitigation usually causes ripple in the next octave or so above cutoff, and throat diffraction, where employed, often causes ripple much higher in frequency. Each of these features are somewhere within the horn, and where closer to the throat, approximately proportionally higher in frequency.
But I must circle back around to add that it isn't just abrupt transitions that cause ripple. Again, some combinations of flare profile, mouth area/length, and front/rear chambers cause a lot of ripple even where the contour appears smooth. You can't tell the response of a device just by casually looking at it. Even the prettiest horns can be highly resonant.
Sure, you can equalize away all the resonances, but my experience is that doesn't solve everything. That approach has never been satisfactory to me. It always sounds artificial, smoothing out resonances with notch filters, compared to using sound sources that are naturally smooth.
I'm not certain what scale is used. Sure, it's not perfect and no speaker is. Yes.To compare speakers and measurements is very difficult.
The wide dispersion can also be a benefit, but it totally depends on the purpose and goal. And like I've previously said, it doesn't take much more sidewall treatment then a waveguide speaker anyway if those reflections want to be attenuated. I know that from comparison and measurements, at least in my rooms. And even more so if the content of the reflections are looked at more closely. We also know vertical reflections are considered only detrimental, while sidewall contribution are more for discussion and taste.
I think one thing is largerly overlooked. That is the freqeuncy response. A speaker with waveguide and CD down to 800-1200 Hz will have an even response in that area and function as a normal speaker below. The CBT36 with controlled directivity, though it's wide horizontally, not only yields an even response above 1 Khz, but also below and more or less down to schroeder. One cannot overlook that. Overall, the frequency response is simply much flatter.
That's possible with a horn with CD lower too but it requires a lot of size or a cornerhorn like Wayne advocates. A speaker with horn in only the highs is a good speaker, but a considerable compromise IMO.
That is simply not true. The vertical directivity is narrower but the response is not "flatter". I "overlook that" because I simply don't see it as an important factor at all. Absorbing the side wall reflections just isn't that simple - it makes the room dead. Better is controlled NARROW directivity in a lively room.
^ The ceiling and floor reflections are the earliest and loudest reflections in most listening rooms:
(Devantier, Audio Engineering Society Convention Paper 5638)
(Devantier, Audio Engineering Society Convention Paper 5638)
Attachments
Last edited:
Hi
Floor and ceiling reflections would make much more of a difference if our ears were vertically oriented or if one were trying to produce a “stereo window” with 4 channels, two upper and two lower right and left speakers.
Side reflections are much more of a detriment because our ears have much more ability to tell left / right and we are trying to produce an image left to right albeit in one plane (as there is no vertical information in the two horizontal channels).
The mono phantom (the ability to produce a strong one with the L ad R speakers not being noticeable) seems to be one reoccurring connection to good stereo imaging regardless of design.
The issue here is how to you preserve as much of the recording as possible when it arrives at the listening position, you do not "get more" preservation when you add room related reflected sound, especially when it conflicts with the dimension of the image we are trying to deliver to the lp..
Best,
Tom
Floor and ceiling reflections would make much more of a difference if our ears were vertically oriented or if one were trying to produce a “stereo window” with 4 channels, two upper and two lower right and left speakers.
Side reflections are much more of a detriment because our ears have much more ability to tell left / right and we are trying to produce an image left to right albeit in one plane (as there is no vertical information in the two horizontal channels).
The mono phantom (the ability to produce a strong one with the L ad R speakers not being noticeable) seems to be one reoccurring connection to good stereo imaging regardless of design.
The issue here is how to you preserve as much of the recording as possible when it arrives at the listening position, you do not "get more" preservation when you add room related reflected sound, especially when it conflicts with the dimension of the image we are trying to deliver to the lp..
Best,
Tom
Yes but in my comparison the PT waveguide had a much smoother response, whereas the H9800 needed a lot of EQ (its off axis response was also much more ragged...). But yet, once everything was equalized the H9800 seemed to sound more "dynamic", whatever that mean...
with all respect Mr Danley, is this not the exact opposite of what the psychoacoustic research says? (project Archimedes, Moulton and Toole, IACC related concert hall literature..).
Actually much of it includes both vertical and horizontal reflections, that neither make much difference past a very short time-period in a small-room context within most standard applications for reproduced sound.
Of course if you place your speakers close to a wall then horizontal reflections become a problem..
As for large-room acoustics (concert hall), that's something else entirely.
Last edited:
Yes, it is true. And I could easily show it. The CBT36 is overall flatter then anything I've tried which must be related to the uniform polar. But again, I know this is possible with a horn as well.
Dealing with sidewall reflections are easier then the ceiling. I care about not only the specular reflections in the highs but all the way down to schroeder. Meaning there's hardly any difference in treatment amount between a waveguide and CBT.
I agree though that using much absorption on sidewalls takes a bit of life out of the music. Therefore I don't use much absorption but angled panels to redirect the energy to the back of the room where I've diffusers. A good replacement for splayed walls.
I both disagree and agree. I diasagree in the respect that while vertical reflections are known to be only detrimental, lateral ones can be preferred. Floyd Toole and the LEDE community both came to this conclusion. And like Markus points out, vertical reflections will also arrive earlier and with a higher level in most small rooms. And let's not forget that the floor is almost impossible to treat in a good matter in a home. A carpet, even a very thick one, will function as an EQ.
If one want a precise image over spaciousness, you might however be right. We're more sensitive to the horizontal.
But bottom line, I don't think it's definite and clear whether what counts the most. It depends on the geoemetry of the room, the speakers and what the listener prefers.
Personally I love the combination of a precise image with early reflections attenuated as much as possible and a highly diffuse soundfield from behind. That's possible to achieve with several kinds of speakers, but a little easier with a horn with CD low in frequency.
Last edited:
- Status
- Not open for further replies.