A good idea. The CBT was designed for concert halls/churches in mind. JBL offers CBT in the pro marked:
JBL :: Product Family
/_MG_2323.jpg)
Omholt's post in message #181 should be clarified as the link he gives for the JBL produced CBT arrays really have no relationship to the photos presented within his post.
The JBL CBT commercial product line is a mass marketed, likely offshore manufactured and produced series of arrays with little direct involvement from Don Keele. These arrays are for modest size venues and likely not intended for use in large auditoriums.
The church photos are from the Marshall Kay "Cary Church of God" project wherein Kay developed a custom solution from Keele's work and personal contact. This project did not involve JBL to my knowledge.
If you go to Don Keele's website you can you can read about this project. Marshall Kay also has additional CBT info on his Audio Artistry website. Marshall and Don developed the Parts Express CBT kit.
Don Keele's CBT (Constant Beamwidth Transducer) Page
The JBL CBT commercial product line is a mass marketed, likely offshore manufactured and produced series of arrays with little direct involvement from Don Keele. These arrays are for modest size venues and likely not intended for use in large auditoriums.
The church photos are from the Marshall Kay "Cary Church of God" project wherein Kay developed a custom solution from Keele's work and personal contact. This project did not involve JBL to my knowledge.
If you go to Don Keele's website you can you can read about this project. Marshall Kay also has additional CBT info on his Audio Artistry website. Marshall and Don developed the Parts Express CBT kit.
Don Keele's CBT (Constant Beamwidth Transducer) Page
Last edited:
Absolutely makes sense to me. Also gets them more up and out of the way. I also seem to remember some psychoacoustic principle that floor reflections tend to sound more natural than ceiling reflections, so if you've gotta have one or the other, that would be yet another plus for ceiling mount.
You have to look at the geometry of your room. Seated or standing, I would guess you are closer to the array center when the array is on the floor than on the ceiling. I have also done measurements of carpet absorption and it is less than you would think. I wouldn't worry about compromises to the performance due to floor mounting, (Also a lot less work).
David
Regarding the array configuration, the manual does cover it in good detail. It is a fairly simple frequency independent weighting similar to the one I previously used in the McIntosh XRT24 (5 zones with 3 weighting levels along the length). The DSP crossover is for crossover between sections and overall EQ, not for achieving the array directivity.
Regards,
David S.
Regards,
David S.
http://www.diyaudio.com/forums/mult...th-transducers-line-arrays-6.html#post2257395
If you want to read about level tapering in line arrays.
David S.
If you want to read about level tapering in line arrays.
David S.
@Jim Griffin
The CBT looks like å great design and I look forward to hear it. But I may go another direction. Do you agree with the points made below (as an answer to my question)?
The CBT looks like å great design and I look forward to hear it. But I may go another direction. Do you agree with the points made below (as an answer to my question)?
Jeno,
Let me address Bear's comments as below:
1. The constant beam width thing as far as I can tell is talking about the VERTICAL dispersion.
Yes, line arrays are all about changing the vertical dispersion.
2. The horizontal dispersion is more or less the same as any line source - except to the extent that it is possible to seriously limit the output of all but a single center driver, and so have an array at HF have the same dispersion as say a single tweeter. That would imply frequency tapering, and not amplitude tapering, if I am correct on this count. Keele uses amplitude tapering only, iirc.
Keele does use vertical tapering but realize that he attains constant beamwidth and flat frequency response over a significant portion of the array's radiation and frequency band via his curved array and weighting of the amplitude. These are two important factors that distinguish his CBT vs. a straight array.
3. Keele's own paper shows that there is no performance difference between his curved array and a straight line source under two conditions:
- you are either short (sit low) or close in (and below the top of the array by a bit)
- the array is floor to ceiling
You need to read Don's AES Convention Paper dated 2010 November 4-7 entitled: "A Performance Ranking of Seven Different Types of Loudspeaker Line Arrays". Don analyzes seven arrays (2 meter height) including an unshaded straight-line array, shaded straight-line, unshaded J-line, unshaded spiral line, unshaded circular arc, a Legendre shaded CBT circular arc, finally a delay curved CBT straight-line array. The performance parameters show plots show beamwidth uniformity, directivity uniformity, vertical soundfield uniformity, polar sidelobe suppresion, smoothness/flatness of off axis frequency response, sound pressure rolloff vs. distance, and near-far pattern uniformity. This is the definitive comparison paper for the various types of arrays.
4. He So, imo the straight vertical array, floor to ceiling is superior to the curved array on three counts:
- it is simpler to build and has equal or better VERTICAL dispersion characteristcs
- it has higher max SPL, and/or lower THD for a given SPL
- it is easier to build and wire
Yes, if limited to those 3 factors you would conclude that a straight array would be superior to the circular CBT. But you need to consider the other factors wherein the circular arc weighted CBT is superior. Don's 2010 paper addresses why a CBT is better than a straight line array.
5. The CBT, as far as I can tell is intended to be able to provide even HF coverage above and near the top of the array in situations where a floor to ceiling speaker is not feasible or practical.
I'll agree that the CBT would achieve more even coverage but again the CBT would also win versus the straight line array for other factors.
I hope that helps.
Jim
Let me address Bear's comments as below:
1. The constant beam width thing as far as I can tell is talking about the VERTICAL dispersion.
Yes, line arrays are all about changing the vertical dispersion.
2. The horizontal dispersion is more or less the same as any line source - except to the extent that it is possible to seriously limit the output of all but a single center driver, and so have an array at HF have the same dispersion as say a single tweeter. That would imply frequency tapering, and not amplitude tapering, if I am correct on this count. Keele uses amplitude tapering only, iirc.
Keele does use vertical tapering but realize that he attains constant beamwidth and flat frequency response over a significant portion of the array's radiation and frequency band via his curved array and weighting of the amplitude. These are two important factors that distinguish his CBT vs. a straight array.
3. Keele's own paper shows that there is no performance difference between his curved array and a straight line source under two conditions:
- you are either short (sit low) or close in (and below the top of the array by a bit)
- the array is floor to ceiling
You need to read Don's AES Convention Paper dated 2010 November 4-7 entitled: "A Performance Ranking of Seven Different Types of Loudspeaker Line Arrays". Don analyzes seven arrays (2 meter height) including an unshaded straight-line array, shaded straight-line, unshaded J-line, unshaded spiral line, unshaded circular arc, a Legendre shaded CBT circular arc, finally a delay curved CBT straight-line array. The performance parameters show plots show beamwidth uniformity, directivity uniformity, vertical soundfield uniformity, polar sidelobe suppresion, smoothness/flatness of off axis frequency response, sound pressure rolloff vs. distance, and near-far pattern uniformity. This is the definitive comparison paper for the various types of arrays.
4. He So, imo the straight vertical array, floor to ceiling is superior to the curved array on three counts:
- it is simpler to build and has equal or better VERTICAL dispersion characteristcs
- it has higher max SPL, and/or lower THD for a given SPL
- it is easier to build and wire
Yes, if limited to those 3 factors you would conclude that a straight array would be superior to the circular CBT. But you need to consider the other factors wherein the circular arc weighted CBT is superior. Don's 2010 paper addresses why a CBT is better than a straight line array.
5. The CBT, as far as I can tell is intended to be able to provide even HF coverage above and near the top of the array in situations where a floor to ceiling speaker is not feasible or practical.
I'll agree that the CBT would achieve more even coverage but again the CBT would also win versus the straight line array for other factors.
I hope that helps.
Jim
Last edited:
Thanks for your reply!
That paper shows clearly that the CBT is superior in all mentioned areas.
But that paper only shows far field analyzis of a free standing line array. I would like to see how a floor to ceiling straight line array performs in the near field compared to the CBT. I guess that under such circumstances the straight line array would better the free standing straight line array in many areas.
Jan
Jan,
In Keele's 2010 paper he does show the smoothness and flatness of off-axis (0, 6 12, 18, 24, 30 degrees) frequency responses for each of the seven arrays he considered. He has plots at 3 m and 18 m distances. For a 2 m high array the 3m distance is essentially near field while 18 m is far field. Again the unshaded straight array exhibits less uniformity vs. both frequency and distance than the shaded circular CBT array.
I suspect that a floor to ceiling straight array would perform slightly better than a 2 m length array but it still would not perform better than the shaded CBT 2 m array. Now many of us can be satisfied with a straight array in the near field but the shaded CBT will be superior throughout the listening space.
Jim
In Keele's 2010 paper he does show the smoothness and flatness of off-axis (0, 6 12, 18, 24, 30 degrees) frequency responses for each of the seven arrays he considered. He has plots at 3 m and 18 m distances. For a 2 m high array the 3m distance is essentially near field while 18 m is far field. Again the unshaded straight array exhibits less uniformity vs. both frequency and distance than the shaded circular CBT array.
I suspect that a floor to ceiling straight array would perform slightly better than a 2 m length array but it still would not perform better than the shaded CBT 2 m array. Now many of us can be satisfied with a straight array in the near field but the shaded CBT will be superior throughout the listening space.
Jim
Last edited:
A good reality check, Jan. It would be a mistake to take a concept (CBT) conceived to address the free-field needs of large venues and glibly suppose it will translate identically to domestic spaces with 8-9' ceilings. Yes, a plane-coupled CBT array might work fine domestically, but it would also be interesting to compare it to a floor-to-ceiling straight, unshaded array as mentioned.
Regarding the question of CBT's horizontal dispersion WRT frequency, it is certainly not constant. More like a mini monitor. In terms of the principles involved, the radiating width of the midwoofers is about 3", so they are ~1/4 wavelength wide at the 1kHz crossover point, but they are mounted on a baffle that is ~1/2 wavelength wide at that frequency, so their midbass horizontal dispersion probably narrows from 180 to somewhat wider than 90 degrees @ 1kHz. Above the crossover point, the 3/4" tweeters radiate into a baffle-defined 90 degrees until about 12kHz when their radiating width becomes significant WRT frequency, and their radiation begins to narrow.
Regarding CBT vs. full-height unshaded array in a domestic setting, the CBT would probably keep SPL more constant with distance, and it would probably require at least 30% fewer drivers than a comparable floor-to-ceiling array in a typical listening room. But it would certainly give up efficiency and SPL headroom in comparison.
Regarding the question of CBT's horizontal dispersion WRT frequency, it is certainly not constant. More like a mini monitor. In terms of the principles involved, the radiating width of the midwoofers is about 3", so they are ~1/4 wavelength wide at the 1kHz crossover point, but they are mounted on a baffle that is ~1/2 wavelength wide at that frequency, so their midbass horizontal dispersion probably narrows from 180 to somewhat wider than 90 degrees @ 1kHz. Above the crossover point, the 3/4" tweeters radiate into a baffle-defined 90 degrees until about 12kHz when their radiating width becomes significant WRT frequency, and their radiation begins to narrow.
Regarding CBT vs. full-height unshaded array in a domestic setting, the CBT would probably keep SPL more constant with distance, and it would probably require at least 30% fewer drivers than a comparable floor-to-ceiling array in a typical listening room. But it would certainly give up efficiency and SPL headroom in comparison.
Last edited:
Concerning horizontal dispersion of a line array: Any two-way line array (straight, curved, whatever) will have horizontal dispersion commeasurate with the individual drivers. What you do in vertical placement of the array is that you have no additional degree of freedom to alter the horizontal dispersion. Hence, you have essentially the inherent horizontal dispersion of the individual drivers.
Now there is a way to alter the horizontal dispersion but at the expense of adding a second line of woofers--another degree of freedom. Think of an array with a line of woofers, a line of tweeters, and another line of woofers. This is an elemental on its side MTM arrangement which does give you a way to impact the horizontal dispersion by creating off axis nulls in the horizontal direction. McIntosh has used this arrangement for several of their line arrays and you could do the some trick with a circular CBT if you wish.
Now there is a way to alter the horizontal dispersion but at the expense of adding a second line of woofers--another degree of freedom. Think of an array with a line of woofers, a line of tweeters, and another line of woofers. This is an elemental on its side MTM arrangement which does give you a way to impact the horizontal dispersion by creating off axis nulls in the horizontal direction. McIntosh has used this arrangement for several of their line arrays and you could do the some trick with a circular CBT if you wish.
Last edited:
For a straight line array there is no effect on horizontal directivity, but this is not true for curved arrays. A curved array has extra width (beyond a single driver's width) when viewed from side angles. Keele shows that as you move off axis both horizontally and vertically the advantage of the curved array diminishes to being more like an uncurved but shaded array. The depth due to curvature would then increase directivity.
For most constructers the simplicity of a shaded straight line array might be a better bet.
David S.
For most constructers the simplicity of a shaded straight line array might be a better bet.
David S.
Keele's horizontal polar plots of circular arrays show some elongation in the horizontal plane. Think of a eye shape vs. a round plot exhibited by straight arrays. As the curvature increases the eye shape pattern elongates in the horizontal plane.
Likely for home applications the horizontal directivity of a curved array isn't a strong design factor to be considered unless you trend toward increased curvature.
Straight line arrays essentially have omnidirectional horizontal polars.
Likely for home applications the horizontal directivity of a curved array isn't a strong design factor to be considered unless you trend toward increased curvature.
Straight line arrays essentially have omnidirectional horizontal polars.
Last edited:
I seem to recall seeing their product listing for a half year or so, but it's never been available when I've looked, always seemingly sold out. I wonder if they've EVER had kits for sale yet.