That may be because it is strictly forbidden to listen to line arrays in
domestic listening rooms.
Domestic use of line arrays may cause auditory hallucinations and
throbbing as well.
In some cases serious and continuing disorientation in time and space
has been reported.
Ahypnia also is seen as an adverse effect by some scientists.
Keep away from the stuff, you don't wanna end like me.
Kind Regards
Last edited:
Okay- the result is there is only one lobe and half the energy is in it. The rest of the energy changes direction of radiation with frequency so there are no side lobes until like 70 degrees off axis and those are 30dB down. The sound outside the main lobe is very diffuse and non directional and also very not flat at all though is it possible to here some of the content. Moving 6 inches, from out of the beam to in the beam, is very much like moving out of a spotlight to in the spotlight. Very muddy and diffuse into crystal clear. Extremely well defined experience.
As a funny note, people walk into the studio (standing) and wonder why the sound is so strange. Many things have been said but my favorite is "these are "sit down" speakers." The look on peoples face when their ears enter the beam- That's worth a lot!
Watching people bob their head into and out of the beam a few times to check what they just heard- hahahaha!For those that are suggesting the drivers were overloaded on the CBT, the array can deliver much higher SPL with a 70Hz crossover point but since a crossover change is required, I opted to leave the crossover at 50Hz so we could simply switch off the woofer for demos without the woofer. Also, some of the overload you were hearing was the crossover, during the fireworks demo and stearman airplane demo, both the mid and high channels were clipping. There was no simple solution, while in Dayton to restructure the gain and even if we did, the 150W/ch ATI amp would have quickly become the limit.
-mk
BTW - I see David Smith posted his article here, this is one of the best line array articles I personally have ever read. If you are interested in line arrays, I hope you read it.
-mk
BTW - I see David Smith posted his article here, this is one of the best line array articles I personally have ever read. If you are interested in line arrays, I hope you read it.
Is this Don? If so, I hope that you would clear something up. You and I both know that the CBT is not constant directivity in the horizontal direction, its directivity must change with frequency. It seems that there is a great deal of confusion on this point.
I see that Dave did not post my "Letter to the Editor" about his paper. I doubt we will see that.
I see that Dave did not post my "Letter to the Editor" about his paper. I doubt we will see that.
I don't think so--wouldn't he be dbk and not mk? Would he claim to have read Don's papers if he was Don and wrote them?
Anyway, I'm enjoying this thread and the great links.
Nope, not Don, just the guy that built the array you see at the beginning of the thread.
However, if you are suggesting the frequency response changes as you move horizontally away from the frontal axis, the small diameter drivers and low 1kHz crossover point to the tweeter ensure that flatness is maintained to very wide angles. We have done some simple tests but will do a very comprehensive analysis soon that compares the actual sound field to Don's simulations. We will post the results on his website.
The speaker has a very interesting property in that the spectrum is very uniform at extreme off axis locations. One can stand next to the speaker with your ear positioned at the end of the array hear a flat spectrum that is no louder than other locations in the room. Those that hear how well this works and understand inverse square law and 1/r relationships think than I am playing some sort of trick on them.
-mk
However, if you are suggesting the frequency response changes as you move horizontally away from the frontal axis, the small diameter drivers and low 1kHz crossover point to the tweeter ensure that flatness is maintained to very wide angles. We have done some simple tests but will do a very comprehensive analysis soon that compares the actual sound field to Don's simulations. We will post the results on his website.
The speaker has a very interesting property in that the spectrum is very uniform at extreme off axis locations. One can stand next to the speaker with your ear positioned at the end of the array hear a flat spectrum that is no louder than other locations in the room. Those that hear how well this works and understand inverse square law and 1/r relationships think than I am playing some sort of trick on them.
-mk
Last edited:
Quite understandable. They put out a huge amount of sound for such small footprint speakers. If I needed slim column speakers, I would be looking very hard at these. For an average listening room they are plenty more than enough.
I do think you need to look into a better sub, tho.
The sub was built the night before the show, quick and dirty. We did not do any crossover optimization nor did we attempt to apply the correct pole/zero equalization to extend the low end but rather used the peaking circuit built into the amp. It is interesting to me that several attendees at the show commented on multiple occasions that the bass was good and in fact very good. One of those who commented, I consider to be an expert on sub-woofers, however, I respect everyone's opinion and several on this board did not like the sub and I think that we could have done better had we taken time to optimize it a little.
-mk
-mk
Last edited:
Hi,
I'm interested if anyone knows any psychoacoustic studies of line arrays in small domestic listening rooms? I have found none so far.
I like to put the room and the human in the picture as well
- Elias
Most studies on the subject aren't so specific as to evaluate a particular speaker in a domestic room. Rather, there have been many good papers on the audibility of reflections versus all the variables that might occur. Toole's book gives a great summary on both what he and Sean Olive found, and also others such as Ando, Devatier, and Bech.
Researchers have studied what level of reflection is just detectable, what level causes image shift, what level cause audible timbre modification and what level is perceived as a distinct echo. Generally these thresholds are measured vs. arrival time. Direction of arrival is also key. The test noise is important as well (speech, pink noise, impulses).
Knowing all this you could then infer the performance of any specific radiation pattern and room.
The most interesting results are about the difference in impression that lateral and vertical reflections gave. Although the detection thresholds were not too far off, Olive and Toole found "the lateral reflection generating much more of a sensation of spaciousness at levels just above threshold, whereas the vertical reflection, on the median plane, was apparent more as an effect on timbre, with a slight amount of high-frequency spatial diffuseness."* Bech found similarly that the room reflection most likely to be audible was the floor bounce.
The conclusion seems to be that narrower vertical dispersion and very wide lateral dispersion are a good thing. They will reduce the coloration effects of floor ceiling bounces while maximising the lateral spaciousness. The CBT array should be very good for this.
Note that the CBT using the floor is not a "floor bounce' per se, but a mirrored reflection of the second half. Tiled or carpeted, the floor is very reflective for low frequencies and more reflective than you think for high frequencies.
I have looked with interest at the CBT for possible theater use. I don't know about the output capibilities, but the evenness of coverage across a seating area would be great.
Regarding uncontrolled lateral directivity, nothing would stop JBL from using CD contours for achieving any desired lateral directivity, or a 2D array as Audiopioneer suggested. These are primarily a commercial sound product and wide lateral/narrow vertical are the desired dispersions (typical house of worship application).
David Smith
*The Detection of Reflections in Typical Rooms
david, (and others)
It seems to me that in the case of the CBT in a real room, or any other speaker with significant levels of HF energy going up and of course depending somewhat on the height (time delay) of the ceiling that there can also be a shift in perceived "center" of the stereo image. More reflection from above at HF tends to make the image shift up in direction in my experience. This is not related to "coloration".
WRT to the floor "reflection" - and yes it is a bounce - my concern is that anything less than a hard flat surface will not produce a textbook result. The reason for that is that there will be both diffusion of the reflection and alteration in the amplitude and frequency response (for example with "carpet"). This means that one will get varied results depending on the nature of the floor.
The strong upward component of the CBT array is of concern to me - notwithstanding the other apparent benefits...
Of course, for all speakers the room dimensions play a role in what is heard. But since we're talking about line arrays and this one in specific I'm focusing on it.
It seems to me that in the case of the CBT in a real room, or any other speaker with significant levels of HF energy going up and of course depending somewhat on the height (time delay) of the ceiling that there can also be a shift in perceived "center" of the stereo image. More reflection from above at HF tends to make the image shift up in direction in my experience. This is not related to "coloration".
WRT to the floor "reflection" - and yes it is a bounce - my concern is that anything less than a hard flat surface will not produce a textbook result. The reason for that is that there will be both diffusion of the reflection and alteration in the amplitude and frequency response (for example with "carpet"). This means that one will get varied results depending on the nature of the floor.
The strong upward component of the CBT array is of concern to me - notwithstanding the other apparent benefits...
Of course, for all speakers the room dimensions play a role in what is heard. But since we're talking about line arrays and this one in specific I'm focusing on it.
Okay- the result is there is only one lobe and half the energy is in it. The rest of the energy changes direction of radiation with frequency so there are no side lobes until like 70 degrees off axis and those are 30dB down. The sound outside the main lobe is very diffuse and non directional and also very not flat at all though is it possible to here some of the content. Moving 6 inches, from out of the beam to in the beam, is very much like moving out of a spotlight to in the spotlight. Very muddy and diffuse into crystal clear. Extremely well defined experience.
As a funny note, people walk into the studio (standing) and wonder why the sound is so strange. Many things have been said but my favorite is "these are "sit down" speakers." The look on peoples face when their ears enter the beam- That's worth a lot!
Very interesting. But while that may seem meaningful to you since you know what it is, I don't!?!
How about a jpeg or diagramatic explanation?
btw, is this a dipole?
_-_-
I don't really understand your remarks as this unit has very curtailed directivity in the upwards direction.
You need to look at the specs for vertical dispersion. Look up the CBT70J on the JBL pro site. At 30 degrees up the unit is down 12dB broadband (800 Hz on up). In fact this is very typical of most line arrays. If you get your ears above the top unit they drop like a stone, hence greatly reducing any ceiling bounce. The CBT just does it in a much more controlled (constant) fashion than most line arrays. Don't let the upwards curvature fool you, rely on the measured data.
Regarding floor bounce, I was making a distinction between a bookshelf speaker on a stand where a reflected source gives two sound paths and comb filtering from the delay of the second source, versus a line source where the second half is created by the reflection of the first. With all the elements of the CBT being reflected, and with the system we want really being the complete doubled up array, then the floor bounce is a very good thing.
"Buy one get one free", as Line Array said.
As to diffusion from carpet, I'll let Don or Audiopioneer address whether there are any issues with carpeted floors. As I mentioned, carpet reflects more HF than you would guess.
I agree regarding the comments from sumaudioguy; "the result is there is only one lobe and half the energy is in it." They have me scratching my head again. Let's see, energy is pressure squared, integrated over area for power....
David
Last edited:
First Time Post from Don Keele
(Note: this is a very long initial post so bear with me!)
Hi, I’m that crazy bushy-haired old guy standing next to that speaker that’s bent over backwards. I’m Don Keele the inventor (“TheRealDonKeele”). The prototype demonstration CBT system, dubbed the CBT36, is a 5ft-tall two-way 36deg circular-arc ground-plane CBT (Constant Beamwidth Transducer) line array loudspeaker utilizing 18 each 3.5” mid-woofers and 72 each 0.72” miniature wide-band loudspeakers.
The system is crossed over at 1 kHz with a 24 dB/octave acoustic Linquitz-Riley crossover and can be used down to 50 or 60 Hz. A sub is required for lower frequencies. The CBT system, based on unclassified military underwater sound research, is broad-band constant directivity/coverage and is specifically designed to operate over a ground plane without harmful floor reflections and provides an extremely even coverage in all directions from distances near to far.
He’s a Very Cool Guy:
Remember that I was awarded an academy award in 2002 for work I did on cinema constant-directivity loudspeaker systems and this makes me a certified audio authority and a genuine “know it all”. So you’d better believe everything I’m going to tell you! I’m also distantly related to Einstein and have had liaisons with Charlize Theron. All this can be checked out on my web site at www.DBKeele.com.
Downloadable CD:
Side note: My website has all the information you ever wanted to know about the CBT arrays (in excruciating detail) in a downloadable 250 MB goodies CD (http://www.xlrtechs.com/dbkeele.com/Don's CBT Goodies Zip Download.html). Included are all five of my AES CBT papers with Power Points, the original three military ASA papers describing the CBT underwater transducers, posters that describe all my prototypes and measurements, very interesting sound-field simulations, copies of my two CBT patents, tone-burst test wave files, and much much more. Lots of neat stuff here!
Loudspeaker Drivers:
The 3.5” driver is the Parts Express ND90-8 full-range unit (stock #290-210) and the miniature driver is the Harman Multimedia Odyssey 2 high-performance miniature full-range transducer. Yes, you heard me right, it’s a multimedia speaker! Before you pooh-pooh my choice of a multimedia speaker in a high-performance high-end system, consider that it’s one of the finest little speakers I’ve ever come across. (It’s not so cheap either!) The CBT design works best when very small speakers are used in the high-frequency range so that beamwidth and coverage is maintained accurately to the highest frequencies. The Odyssey 2 driver can be placed very close at CC spacings in the range of 0.75” to 0.8”!
The Odyssey 2 is a very-fine wide-range mini speaker with a reasonably-flat response from 500 Hz to 20 kHz. It has a 0.5”-diameter inverted-dome aluminum diaphragm with a free-air resonance of about 500 Hz (very low for a small driver) and a sensitivity of about 75 dB SPL 1W/1m. It also has a 1.5W power rating with a peak linear excursion of 1.4mm and a mechanical-limit excursion of 1.9mm (extremely generous for a driver this size!).
Tone-Burst Tweeter Tests:
I’ve done 6.5-cycle shaped tone-burst peak SPL/power tests on this speaker and found that a single unit above 2 kHz can generate clean bursts with peak SPLs of about 110 dB at 1 m! This burst test is the same that I accomplished on all the loudspeakers I reviewed for Audio Magazine between 1990 and 2000. When you array 72 of these units, it will play extremely loud and clean with any amount of power you care to drive it with!
How Does It Sound?:
Of course my speaker sounds like an AM radio. Many thanks go to Earl Geddes for pointing this out. However, …………. It sounds like the best darn HD wide-band AM radio you ever heard with wide-band constant directivity/beamwidth behavior, extremely even coverage at all points right-left, up-down, and near-far, and can play incredibly loud and clean! All this over a perfectly reflective floor or ground plane. No deleterious floor-bounce effects! It has the most even and uniform coverage of any speaker that has ever existed! (Take that Earl Geddes!)
Two Nagging Forum Questions:
Now for a couple of items that came up on the forum concerning:
1. Operation of a ground-plane CBT array over a carpeted floor, and
2. The horizontal off-axis response
Operation over Carpeted Floor:
One big question that can be raised is what happens when a ground-plane CBT line array, that is designed to operate over an acoustically -reflective floor, is used on a carpeted floor. Quick answer, not a whole lot. Surprisingly, the sound field about four to six inches above the floor is not affected to any great extent at all. However, at points near and on the floor, the high frequencies are rolled off because the high frequencies are radiated across the floor at a low grazing angle where the carpet absorption is at a maximum. Of course, the effects depend on the nature of the carpet, thickness, type etc. In general, carpeting is not an effective absorber except at high frequencies.
Horizontal Off-Axis Response:
The vertical coverage (beamwidth) of a circular-arc array is a strong function of the horizontal off-axis angle and follows a cosine law. It is maximum at straight ahead and gets narrower as you go off axis reaching a minimum at 90 deg off axis. Note that for a circular-arc ground-plane CBT array, the 90 deg minimum vertical angle is actually aimed at floor level, i.e. if you listen to a ground-plane array at 90 deg off axis and gradually squat down, the level gets louder and louder and is loudest at floor level.
This behavior is actually a plus for the circular-arc CBT array, because the side walls are not being sprayed with a lot of energy as compared to a typical straight line array which is mostly omni-directional horizontally. This behavior is fully documented and explained in my third CBT paper titled “Full-Sphere Sound Field of Constant-Beamwidth Transducer (CBT) Loudspeaker Line Arrays” available from my web site (http://www.xlrtechs.com/dbkeele.com/PDF2/Keele (200303 AES Preprint) - CBT Paper3.pdf).
Thanks for listening!
(Note: this is a very long initial post so bear with me!)
Hi, I’m that crazy bushy-haired old guy standing next to that speaker that’s bent over backwards. I’m Don Keele the inventor (“TheRealDonKeele”). The prototype demonstration CBT system, dubbed the CBT36, is a 5ft-tall two-way 36deg circular-arc ground-plane CBT (Constant Beamwidth Transducer) line array loudspeaker utilizing 18 each 3.5” mid-woofers and 72 each 0.72” miniature wide-band loudspeakers.
The system is crossed over at 1 kHz with a 24 dB/octave acoustic Linquitz-Riley crossover and can be used down to 50 or 60 Hz. A sub is required for lower frequencies. The CBT system, based on unclassified military underwater sound research, is broad-band constant directivity/coverage and is specifically designed to operate over a ground plane without harmful floor reflections and provides an extremely even coverage in all directions from distances near to far.
He’s a Very Cool Guy:
Remember that I was awarded an academy award in 2002 for work I did on cinema constant-directivity loudspeaker systems and this makes me a certified audio authority and a genuine “know it all”. So you’d better believe everything I’m going to tell you! I’m also distantly related to Einstein and have had liaisons with Charlize Theron. All this can be checked out on my web site at www.DBKeele.com.
Downloadable CD:
Side note: My website has all the information you ever wanted to know about the CBT arrays (in excruciating detail) in a downloadable 250 MB goodies CD (http://www.xlrtechs.com/dbkeele.com/Don's CBT Goodies Zip Download.html). Included are all five of my AES CBT papers with Power Points, the original three military ASA papers describing the CBT underwater transducers, posters that describe all my prototypes and measurements, very interesting sound-field simulations, copies of my two CBT patents, tone-burst test wave files, and much much more. Lots of neat stuff here!
Loudspeaker Drivers:
The 3.5” driver is the Parts Express ND90-8 full-range unit (stock #290-210) and the miniature driver is the Harman Multimedia Odyssey 2 high-performance miniature full-range transducer. Yes, you heard me right, it’s a multimedia speaker! Before you pooh-pooh my choice of a multimedia speaker in a high-performance high-end system, consider that it’s one of the finest little speakers I’ve ever come across. (It’s not so cheap either!) The CBT design works best when very small speakers are used in the high-frequency range so that beamwidth and coverage is maintained accurately to the highest frequencies. The Odyssey 2 driver can be placed very close at CC spacings in the range of 0.75” to 0.8”!
The Odyssey 2 is a very-fine wide-range mini speaker with a reasonably-flat response from 500 Hz to 20 kHz. It has a 0.5”-diameter inverted-dome aluminum diaphragm with a free-air resonance of about 500 Hz (very low for a small driver) and a sensitivity of about 75 dB SPL 1W/1m. It also has a 1.5W power rating with a peak linear excursion of 1.4mm and a mechanical-limit excursion of 1.9mm (extremely generous for a driver this size!).
Tone-Burst Tweeter Tests:
I’ve done 6.5-cycle shaped tone-burst peak SPL/power tests on this speaker and found that a single unit above 2 kHz can generate clean bursts with peak SPLs of about 110 dB at 1 m! This burst test is the same that I accomplished on all the loudspeakers I reviewed for Audio Magazine between 1990 and 2000. When you array 72 of these units, it will play extremely loud and clean with any amount of power you care to drive it with!
How Does It Sound?:
Of course my speaker sounds like an AM radio. Many thanks go to Earl Geddes for pointing this out. However, …………. It sounds like the best darn HD wide-band AM radio you ever heard with wide-band constant directivity/beamwidth behavior, extremely even coverage at all points right-left, up-down, and near-far, and can play incredibly loud and clean! All this over a perfectly reflective floor or ground plane. No deleterious floor-bounce effects! It has the most even and uniform coverage of any speaker that has ever existed! (Take that Earl Geddes!)
Two Nagging Forum Questions:
Now for a couple of items that came up on the forum concerning:
1. Operation of a ground-plane CBT array over a carpeted floor, and
2. The horizontal off-axis response
Operation over Carpeted Floor:
One big question that can be raised is what happens when a ground-plane CBT line array, that is designed to operate over an acoustically -reflective floor, is used on a carpeted floor. Quick answer, not a whole lot. Surprisingly, the sound field about four to six inches above the floor is not affected to any great extent at all. However, at points near and on the floor, the high frequencies are rolled off because the high frequencies are radiated across the floor at a low grazing angle where the carpet absorption is at a maximum. Of course, the effects depend on the nature of the carpet, thickness, type etc. In general, carpeting is not an effective absorber except at high frequencies.
Horizontal Off-Axis Response:
The vertical coverage (beamwidth) of a circular-arc array is a strong function of the horizontal off-axis angle and follows a cosine law. It is maximum at straight ahead and gets narrower as you go off axis reaching a minimum at 90 deg off axis. Note that for a circular-arc ground-plane CBT array, the 90 deg minimum vertical angle is actually aimed at floor level, i.e. if you listen to a ground-plane array at 90 deg off axis and gradually squat down, the level gets louder and louder and is loudest at floor level.
This behavior is actually a plus for the circular-arc CBT array, because the side walls are not being sprayed with a lot of energy as compared to a typical straight line array which is mostly omni-directional horizontally. This behavior is fully documented and explained in my third CBT paper titled “Full-Sphere Sound Field of Constant-Beamwidth Transducer (CBT) Loudspeaker Line Arrays” available from my web site (http://www.xlrtechs.com/dbkeele.com/PDF2/Keele (200303 AES Preprint) - CBT Paper3.pdf).
Thanks for listening!
Don has measured earlier GPCBT prototypes over carpet and found that they work very well, but since this question is appearing often, we plan to measure the speaker on a hard flat surface and then insert a large piece of carpet in front of the speaker and repeat the measurement. We will post these results also.
-mk
-mk
Last edited: