I didn't like the cbts when I heard them.
I wonder if the improvement you heard was a mental adjustment to the unique radiation pattern?
For instance, it took a couple days for me to adjust to the sound of the Gedlee Summas. At first they sounded too dull. (As you noted with your CBTs.)
Nowadays I can't listen to dome tweeters; they sound too bright. I've adjusted to preferring constant directivity.
At CES, at the Parts Express booth. This was 2013 iirc. The sound was diffuse and undefined, sounded like a Bose Wave Radio. Not that there's anything wrong with a Bose Wave Radio; but the CBT seemed like an awful lot of trouble to create a sound that you can get from $20 worth of paper coned woofers and a transmission line.
Abbey and CBT36 in a small home theater room.
I have measured both speakers in several rooms with several placements. In every case the CBT36 has been extremely flat down to 300-400 Hz and well behaved lower as well. Some times the freq. response has been within +/-3 down between 300 Hz and 20KHz with 1/24 oct. smoothing.
The Abbey is very flat in the range of the waveguide, meaning down till around 1000 Hz. Below that the freq. response becomes uneven and it is nothing like the CBT.
Looking at the reflections (impulse/ETC), the CBT36 has in most cases less early reflections. It completely avoids the floor reflections and ceiling reflections are greatly attenuated. Abbey basically minimizes the reflections from the sidewalls. However, Abbey only does that above 1000 Hz and this can be seen in a bandlimited ETC. The result is, in my opinion, that Abbey is actually more dependent on sidewall treatments than CBT. Why? Because the reflected energy varies, thus coloring the sound.
CBT on the other hand has a wide but uniform response down to 300 Hz. So while you obviously get sidewall reflections, the reflected energy resembles the direct sound much more and it sounds natural and spacious without treatment. Flutter-echo needs to be dealt with for a proper result and perhaps that's what Patrick heard when listening to them. The room shouldn't be too reflective. And if one desires the best imaging, side wall treatment is needed like the Abbey also need. Nothing different there really. Avoiding reflections only above 1000 Hz will yield a great result but actually coloring the sound. So both need broadband sidewall treatment for accuracy and pin pointing.
I have had couple of friends listened to both speakers. They have preferred CBT36 by a large margin. There's isn't much comparison really. It's not only about the frequency response, though that's important. The CBT36 despite it's low sensitivity has way more energy and resembles live sound to a larger degree. The Abbey is lacking in clarity, openness and energy compared to the CBT.
While CBT36 can work decently without a subwoofer both speakers definitely needs subwoofer(s) for the best result. While I like the Abbeys and think it's a well design and above most commercial speakers, the CBT36 is pretty amazing. Especially considering the price, and the low cost tweeter drivers with low sensitivity. Wondering what a CBT with much better drivers will sound like! Thank you Don for an incredible design.
I have measured both speakers in several rooms with several placements. In every case the CBT36 has been extremely flat down to 300-400 Hz and well behaved lower as well. Some times the freq. response has been within +/-3 down between 300 Hz and 20KHz with 1/24 oct. smoothing.
The Abbey is very flat in the range of the waveguide, meaning down till around 1000 Hz. Below that the freq. response becomes uneven and it is nothing like the CBT.
Looking at the reflections (impulse/ETC), the CBT36 has in most cases less early reflections. It completely avoids the floor reflections and ceiling reflections are greatly attenuated. Abbey basically minimizes the reflections from the sidewalls. However, Abbey only does that above 1000 Hz and this can be seen in a bandlimited ETC. The result is, in my opinion, that Abbey is actually more dependent on sidewall treatments than CBT. Why? Because the reflected energy varies, thus coloring the sound.
CBT on the other hand has a wide but uniform response down to 300 Hz. So while you obviously get sidewall reflections, the reflected energy resembles the direct sound much more and it sounds natural and spacious without treatment. Flutter-echo needs to be dealt with for a proper result and perhaps that's what Patrick heard when listening to them. The room shouldn't be too reflective. And if one desires the best imaging, side wall treatment is needed like the Abbey also need. Nothing different there really. Avoiding reflections only above 1000 Hz will yield a great result but actually coloring the sound. So both need broadband sidewall treatment for accuracy and pin pointing.
I have had couple of friends listened to both speakers. They have preferred CBT36 by a large margin. There's isn't much comparison really. It's not only about the frequency response, though that's important. The CBT36 despite it's low sensitivity has way more energy and resembles live sound to a larger degree. The Abbey is lacking in clarity, openness and energy compared to the CBT.
While CBT36 can work decently without a subwoofer both speakers definitely needs subwoofer(s) for the best result. While I like the Abbeys and think it's a well design and above most commercial speakers, the CBT36 is pretty amazing. Especially considering the price, and the low cost tweeter drivers with low sensitivity. Wondering what a CBT with much better drivers will sound like! Thank you Don for an incredible design.
Attachments
Be sure to listen to the Abbey at the proper listening height. I didn't measure my Summas until two days before I sold them, and it was only then that I discovered my listening height was all wrong. Due to the very large CTC gap between woofer and tweeter, the listening height must be *exactly* right with this speaker. Get it wrong by even three inches and the midrange has a huge suckout.
This is because of pathlength. For instance, the tweeter in the Summa is over forty centimeters away from the midrange. The crossover point is 900hz. Due to this huge gap, if your seating position changes the pathlength by even 10cm (4") then the phase of the midrange and the tweeter will be one quarter wavelength out of phase. Get the height wrong by 8", which is easy to do, and you have a humongous null, right in the midrange.
Conventional speakers don't suffer from this in such a huge degree; if your tweeter is 10-15cm away from your midrange you have more flexibility with the vertical height.
Another contributing factor here is that speakers with a 15" waveguide are typically used for prosound. If you're listening to a speaker like this in a prosound venue and you're standing twenty feet away, the vertical beamwidth is less of a problem.
Does that make sense? You might have a vertical window that's only fifteen degrees wide, but if you're standing twenty feet away, that margin of error is much wider, because you're so far.
If you read some of my early reviews of the Summa you'll note that I preferred listening to it far away. At the time I could only speculate on why this is, but after measuring I know why: it's very picky about seat height.
This is because of pathlength. For instance, the tweeter in the Summa is over forty centimeters away from the midrange. The crossover point is 900hz. Due to this huge gap, if your seating position changes the pathlength by even 10cm (4") then the phase of the midrange and the tweeter will be one quarter wavelength out of phase. Get the height wrong by 8", which is easy to do, and you have a humongous null, right in the midrange.
Conventional speakers don't suffer from this in such a huge degree; if your tweeter is 10-15cm away from your midrange you have more flexibility with the vertical height.
Another contributing factor here is that speakers with a 15" waveguide are typically used for prosound. If you're listening to a speaker like this in a prosound venue and you're standing twenty feet away, the vertical beamwidth is less of a problem.
Does that make sense? You might have a vertical window that's only fifteen degrees wide, but if you're standing twenty feet away, that margin of error is much wider, because you're so far.
If you read some of my early reviews of the Summa you'll note that I preferred listening to it far away. At the time I could only speculate on why this is, but after measuring I know why: it's very picky about seat height.
Mine are placed at listening height. What audiophile wouldn't? I've also measured them at slightly different heights (all within normal listening height) without seeing any noticeable difference.
With the CBTs though, the height basically means nothing.
With the CBTs though, the height basically means nothing.
Attachments
I built floor-to-ceiling arrays and they are documented here:
http://www.diyaudio.com/forums/multi-way/284371-corner-floor-ceiling-line-array-using-vifa-tc9.html
Shading and delay are required to maintain vertical uniformity when the line is not "infinite." I made in room measurements to prove the vertical uniformity and also dispel the notion of comb filtering with line arrays. Posts 127-130:
http://www.diyaudio.com/forums/mult...line-array-using-vifa-tc9-13.html#post4584857
I heard the CBTs and LX521 at BAF also. I was not impressed with the CBTs. The LX521 sounded much better tonally. But a lot of it is down to the crossover and overall response. There is no doubt in my mind that arrays tackle the right problems in the home environment, i.e., floor and ceiling reflections. These reflections color the sound. Side reflections add spaciousness. Side reflections are also vastly simpler to treat.
With my arrays, the sound is much cleaner, much more intelligible, the music is so much easier to follow. One of the startling things about these arrays is that you can hear all the music very clearly right until the point it becomes inaudible. It's not like you can turn it down and not pay attention. It commands your attention even at super low volumes. I think this is because the mind is being fed the right cues at the right times and is not having to do a lot of work.
Anyway, in summary, I think arrays, especially floor-to-ceiling arrays are the right solution for the home environment. Small, even tiny footprint, quite good efficiency, high SPL capability, low distortion, and most important, avoidance of floor and ceiling reflections.
http://www.diyaudio.com/forums/multi-way/284371-corner-floor-ceiling-line-array-using-vifa-tc9.html
Shading and delay are required to maintain vertical uniformity when the line is not "infinite." I made in room measurements to prove the vertical uniformity and also dispel the notion of comb filtering with line arrays. Posts 127-130:
http://www.diyaudio.com/forums/mult...line-array-using-vifa-tc9-13.html#post4584857
I heard the CBTs and LX521 at BAF also. I was not impressed with the CBTs. The LX521 sounded much better tonally. But a lot of it is down to the crossover and overall response. There is no doubt in my mind that arrays tackle the right problems in the home environment, i.e., floor and ceiling reflections. These reflections color the sound. Side reflections add spaciousness. Side reflections are also vastly simpler to treat.
With my arrays, the sound is much cleaner, much more intelligible, the music is so much easier to follow. One of the startling things about these arrays is that you can hear all the music very clearly right until the point it becomes inaudible. It's not like you can turn it down and not pay attention. It commands your attention even at super low volumes. I think this is because the mind is being fed the right cues at the right times and is not having to do a lot of work.
Anyway, in summary, I think arrays, especially floor-to-ceiling arrays are the right solution for the home environment. Small, even tiny footprint, quite good efficiency, high SPL capability, low distortion, and most important, avoidance of floor and ceiling reflections.
Thanks for your feedback. I think some of the issue is how the Behringer DSP was set up. I also took a different approach with my CBT which I think most listeners would find to be quite a bit better than the CBT36 (though it should be given the much higher cost in drivers). The implementation is also different with what I built and that might factor in to what you noticed lacking in the CBT36.
Your measurements shows clearly combing at high frequency. I did a overlay with CBT36 at the approximately same distance and height (12 feet distance and 37" height). Obviously the room acoustics will have an impact, but I still think it shows the problem of using 3,5" drivers in a line array.
Red is CBT36 is orange your line array. Showing it with 1/48 smoothing.
Attachments
There is no point in looking at the unsmoothed responses for HF. You are admitting all the reflections. FWIW, even the red line looks to have some kind of ripple beginning at 3 kHz.
The point is the farther you sit, the higher in frequency comb filtering begins. From where I'm sitting, it doesn't start until 16 kHz. Of course, a line of 1" tweeters will perform better. But is it audibly better? I doubt it. I am super happy with the HF performance of these arrays. I did consider the Tymphany PMT line. Check it out. But they would probably have trouble reaching 100 Hz, even 75 of them.
Anyway, this is off topic for this thread. Why don't you post in my thread if you are interested in discussing further? I would love to see similar height/distance measurements done on the CBT in a real room. Not because I want to point out warts, but because I want to study the implications of a full line verses a shortened, albeit, delayed and shaded, line.
My original point was that the CBT should be superior to the LX521 at least in terms of theory. The rest is execution. Clearly, SL is highly skilled in the execution. I loved the sound of the LX521.
The point is the farther you sit, the higher in frequency comb filtering begins. From where I'm sitting, it doesn't start until 16 kHz. Of course, a line of 1" tweeters will perform better. But is it audibly better? I doubt it. I am super happy with the HF performance of these arrays. I did consider the Tymphany PMT line. Check it out. But they would probably have trouble reaching 100 Hz, even 75 of them.
Anyway, this is off topic for this thread. Why don't you post in my thread if you are interested in discussing further? I would love to see similar height/distance measurements done on the CBT in a real room. Not because I want to point out warts, but because I want to study the implications of a full line verses a shortened, albeit, delayed and shaded, line.
My original point was that the CBT should be superior to the LX521 at least in terms of theory. The rest is execution. Clearly, SL is highly skilled in the execution. I loved the sound of the LX521.
Last edited:
Window for hearing is very short at high frequencies. You can try to gate it to get an idea of what we will hear. But where do you set the gate? 5 ms, 7 ms, 10 ms? In reality, many think that it is a continuously sliding window. REW has introduced this continuous sliding window. It's under the IR windows panel. You can set it to 5 cycles, i.e., admit 5 cycles at 10 kHz and at 100 Hz. I think this is how we hear. I also really like the ERB smoothing in REW because it correlates well with what I hear. I don't hear the large peaks and dips at 10 kHz. I have owned full range drivers with 10 db peaks at 10 kHz before and these arrays sound nothing like that.
If you have an anechoic chamber, please by all means show us the unsmoothed response. Even then, you can ignore what happens after a few ms for the HF. You just don't hear it.
If you have an anechoic chamber, please by all means show us the unsmoothed response. Even then, you can ignore what happens after a few ms for the HF. You just don't hear it.
Omholt. Can you share some of your measurements of both Abbeys and CBT36 in the same room? I own the Abbeys and I do like them very much. In pursuing the white rabbit I want to check some new designs but I must be sure to exchange for something better not the vice versa. The only designs I consider are CBT36 and PSE 144. I like LX521 but they need more space around than I can offer.
May 10, 2016 Don and Kevin Keele have released "The CBT Chronicles" on 'The Official D.B. Keele, Jr. Youtube page'. The CBT Chronicles are a nine-part video series by inventor Don Keele covering the history and technology of CBT (Constant Beamwidth Transducer) loudspeaker arrays including detailed comparative measurements of a traditional speaker system and a CBT array. You can access the playlist at: http://tinyurl.com/CBTChronicles (redirecting to Youtube)
For further information contact Don Keele at DKeeleJr@Comcast.net
For further information contact Don Keele at DKeeleJr@Comcast.net
What is the right height for listening then? Midpoint between woofer and tweeter or on-axis with a tweeter? I found posts with both recommendations and I am not sure which is technicaly true.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.