cannot be done passively
A curved array only provides fixed delay between drivers. As you say parallel/series gives amplitude only "shading." My array uses specific filters to derive transfer functions for each driver. I use an amplifier for each driver to control the driver well. The result is then one main lobe with half the radiated energy in that lobe. The other half of the energy goes all over the place and is without directivity or phase. This particular setup has constant vertical directivity from 200Hz up with an included angle of 12 degrees, a DI of 10. The horizontal is controlled by the drivers themselves with a constant included angle of radiation of 60 degrees. It is extremely apparent when moving from the beam to out of the beam, extremely apparent. This cannot be accomplished by bending the line or by passive mechanical treatments. Obviously, constant directivity equalization is required throughout this pass band.
A curved array only provides fixed delay between drivers. As you say parallel/series gives amplitude only "shading." My array uses specific filters to derive transfer functions for each driver. I use an amplifier for each driver to control the driver well. The result is then one main lobe with half the radiated energy in that lobe. The other half of the energy goes all over the place and is without directivity or phase. This particular setup has constant vertical directivity from 200Hz up with an included angle of 12 degrees, a DI of 10. The horizontal is controlled by the drivers themselves with a constant included angle of radiation of 60 degrees. It is extremely apparent when moving from the beam to out of the beam, extremely apparent. This cannot be accomplished by bending the line or by passive mechanical treatments. Obviously, constant directivity equalization is required throughout this pass band.
I worked on arrays at McIntosh and had very useful improvements of directivity with straight lines and shading techniques. If you add frequency tapering you can do even better.
Certainly an active approach or the use of delay will give you mare variables to play with but don't discount the simpler approaches.
David
Certainly an active approach or the use of delay will give you mare variables to play with but don't discount the simpler approaches.
David
In my line array i use a combination of weighting by
driver distance and frequency dependent power tapering
for the vertical control of radiation angle.
The horizontal angle is controled by the witdh of the
open baffle relative to the drivers.
pics show the input voltage of the lower and the upper
driver trio.
Tweeters on the rear are used to compensate
falling back radiation with rising frequency
of the fullrage drivers, they move in antiphase
to maintain dipole operation >3Khz.
All compensation is done passively with very low
phase shift and smooth slope.
lower frequency limit 60-70 Hz depending on
room and position, dipole subwoofer needed.
cutoff to subwoofer -6db /octave at 80 Hz using
PLLXO.
driver distance and frequency dependent power tapering
for the vertical control of radiation angle.
The horizontal angle is controled by the witdh of the
open baffle relative to the drivers.
pics show the input voltage of the lower and the upper
driver trio.
Tweeters on the rear are used to compensate
falling back radiation with rising frequency
of the fullrage drivers, they move in antiphase
to maintain dipole operation >3Khz.
All compensation is done passively with very low
phase shift and smooth slope.
lower frequency limit 60-70 Hz depending on
room and position, dipole subwoofer needed.
cutoff to subwoofer -6db /octave at 80 Hz using
PLLXO.
Attachments
Last edited:
On his Linkwitz Lab website Dr. Linkwitz has some interesting things to say about the Constant Beamwidth Transducer (CBT) technology. He also includes links to Don Keele's Boston AES CBT presentation earlier this year. I suggest that those interested review Keele's presentation viewgraphs and the movie of his pitch. Linkwitz concludes: "It will be the end of the horn and compression driver in quality critical applications."
Linkwitz-Links
Linkwitz-Links
Would never discount simple approaches myself however, fancy approaches can achieve fancier results. The simply approach is used widely now for concert setup and almost every big player, Peavey, EV, McCormick, etc., has arrays of the simple kind.
Jim Griffin's mention and going to Linkwitz-Links showed a huge amount of useful information- very cool link, thanks.
As a note, the signal to noise ratio and lack of distortion of the amplifier for every driver array is really astounding. With volume set on "blast" I cannot even see the 6.5" midrange drivers move and that is high pass crossover at 100Hz at -12dB per octave.
Jim Griffin's mention and going to Linkwitz-Links showed a huge amount of useful information- very cool link, thanks.
As a note, the signal to noise ratio and lack of distortion of the amplifier for every driver array is really astounding. With volume set on "blast" I cannot even see the 6.5" midrange drivers move and that is high pass crossover at 100Hz at -12dB per octave.
I'll take that bet. Well "horns" are already dead, but the waveguide era is not going away any time soon.
Reading through some of the information, this approach seems somewhat familiar. Makes me think of some of Ted Jordan's work back in th 80's. I think this is a valid approach for large audiances where high SPL is required. Trying to understand how and what different people listen for, aside from tone balance, it seems that there is a group that listens for dynamics and impact, while a different group will listen for the small details.
The people that hear the details take pleasure in the playing techniques and fingerwork of the player. I have heard people complain that they cannot hear this in most of the so called hi-fi systems.
The people that listen to the overall dynamics and impact might complian it's hard to find a system that delivers such. This is the group that the line array will be appealing to.
To learn how Constant Beamwidth speakers really work please visit Don's updated and expanded WEB site (www.dbkeele.com). There you will find all his papers on CBT's. In addition there are test signals and a whole lot more.
I have built several different versions of the CBT and if implemented properly, they behave exactly as describe in these papers and they can compete with any high end system available.
I have built several different versions of the CBT and if implemented properly, they behave exactly as describe in these papers and they can compete with any high end system available.
Well, I heard Don's arrays. While they may not be the worlds best speaker, they certainly do NOT sound like an AM radio. Even without the sub. And as far as I could tell, everything Don illustrated in his directivity plots, he proved with this speaker. The only obvious problems I heard where that he tended to overdrive them, and his sub was boomy. Can't blame the sub boom on the array!
It's fun to play armchair general and criticize what you haven't heard. Once you've heard it, you might not be so quick to dismiss it.
I've also spent time with the ClairAudient line arrays. No subwoofer there and plenty of bass from those 3" drivers. Not an AM radio sound at all. Not perfect, either, but better than many other speakers.
Jim G can chime in. He was in Dayton too and heard the Keele array.
It's fun to play armchair general and criticize what you haven't heard. Once you've heard it, you might not be so quick to dismiss it.
I've also spent time with the ClairAudient line arrays. No subwoofer there and plenty of bass from those 3" drivers. Not an AM radio sound at all. Not perfect, either, but better than many other speakers.
Jim G can chime in. He was in Dayton too and heard the Keele array.
Hey Guys,
you might be interested in this approach.
MLA? Multi-cellular Loudspeaker Array. Martin Audio Ltd.
you might be interested in this approach.
MLA? Multi-cellular Loudspeaker Array. Martin Audio Ltd.
Have you heard Keele's? If you have, it might be apropos to describe the sound. If not you should have stayed silent. This kind of self-interested bashing and silly generalizations doesn't serve anyone.
Last edited:
Don Keele is an old friend and I have talked with him about this not only lately but long ago when he first came across the principles involved. I have not heard Don's actual array, and I didn't say that I did (the topic of this thread is not CBT, but line arrays in general, my comments were not aimed specifically at the CBT), but I have heard many small speaker arrays and I will stand by my comments. Don is well aware of the limitations of his design (the horizontal radiation pattern is not controlled).
"Describe the sound" is not something that I do very often, if you read my work you should know that, I base my claims on technical data and analysis and I probably should have left off the subjective comments.
The fact is that Line arrays are not Constant Beam Width and can never be (unless laid on their side, which is not what they are designed to do). Sound Quality wise, maybe Don's is the exception, but I don't see how it can be since it is not really all that different than conventional line arrays and has to live within the physical constraints imposed by the design.
"Describe the sound" is not something that I do very often, if you read my work you should know that, I base my claims on technical data and analysis and I probably should have left off the subjective comments.
The fact is that Line arrays are not Constant Beam Width and can never be (unless laid on their side, which is not what they are designed to do). Sound Quality wise, maybe Don's is the exception, but I don't see how it can be since it is not really all that different than conventional line arrays and has to live within the physical constraints imposed by the design.
Nobody said you were bashing Keele. Just because a design doesn't have horizontal pattern control doesn't make it bad, and many small inexpensive drivers doesn't mean it will sound poorly. The devil is in the details.
If you base your claims on technical data, then do so. I admired your Bandpass articles in '89 when I read them. I've read your HT book, and my honest opinion of that work is you need an editor and proofreader very badly - again, details
I'm not quite clear on this or why it's a problem. Don's CBT seemed to have remarkable consistency in the vertical axis, both near and far. I spent a lot of time walking around the speakers and the room listening critically. No peaks or valleys that I could hear - not up close or far away. That would be from 2 feet to ~100 feet.
They also sounded very consistent on the horizontal axis, but they were not omni. At the center of the array (the floor in this case) there was a fairly wide pattern which tapered off to very tight at the top. This fascinated me, as I'd never heard that pattern before. Mind you, this did not sound like frequency variation, just amplitude.
The sound was very consistent around the room, at least in front of the speakers. The tonal balance in the room and down the hall was quite similar.
So my experience was that horizontal power response was very consistent in front of the speakers. Amplitude was lower in the rear, especially at the end of the array, but this was not noticeable from in front. Varying from maybe cardioid to hyper-cardioid. FR did not vary enough in either axis to really notice.
Seems the horizontal pattern should be primarily determined by the crossover and driver spacing. Like the graphs SUM has shown or the http://www.falstad.com/interference/index_ap.html interference applet.
Dan
Dan
Anecdotal evidence aside, they are designed for a tight control over the vertical pattern and I have no doubt that they do that well. But horizontally they will not hold a pattern width just as any other source flat source of the same width. The vertical control does not apply to the horizontal and nothing is done to control the horizontal coverage. Look at the application that they were designed for in Toole's book. Vertical control was all that was desired.
To what extent this (horizontal beam width) is important depends on who you are talking to, but the CBT does not fit into the description of this thread because it is not "constant beam width". It is constant beam height. Lets at least get the facts straight.
There is no dictionary definition of beamwidth in accordance to sound that I can find with a quick search, but there is plenty of uses of the term in radio and light...
Telecommunication Standard Terms
1. In the radio regime, of an antenna pattern, the angle between the half-power (3-dB ) points of the main lobe , when referenced to the peak effective radiated power of the main lobe. (188 ) Note: Beamwidth is usually expressed in degrees. It is usually expressed for the horizontal plane, but may also be expressed for the vertical plane. 2. For the optical regime, see beam divergence .
Beamwidth definition by Babylon's free dictionary
I really suggest to you to listen to these systems. I drove all the way from Central Florida recently to hear these and help videotape comparisons to traditional systems because my father's enthusiasm. You mention 5 years and no product? The time is now because of recent advances is miniature full-range drivers that have been ushered into the industry by a need derived by Iphones, laptops, etc. Don's not using AM radio speakers!
I've heard Keele's CBT and recognize them as having a secure place in sound reproduction.
Let's set the bose-style marketing aside and cut to the chase.
Let's get Keele's CBT next to Geddes' waveguides and have a listen, shall we?
Fun for all! ...warts and beauty marks on full display.
.click. rant off
Let's set the bose-style marketing aside and cut to the chase.
Let's get Keele's CBT next to Geddes' waveguides and have a listen, shall we?
Fun for all! ...warts and beauty marks on full display.
.click. rant off
Last edited:
Well, my 2 cents worth - I'm not sure that vertical beamwidth control is that valuable if one has a tall line source to begin with??
As far as Dr. Geddes earlier comment, I definitely agree that a line array is no better than the quality of a single driver. It's just going to play much louder for max SPL and lower in distortion for a given SPL level than a single driver... so pick a darn good sounding driver if you want to make a line array (my advice).
_-_-bear
As far as Dr. Geddes earlier comment, I definitely agree that a line array is no better than the quality of a single driver. It's just going to play much louder for max SPL and lower in distortion for a given SPL level than a single driver... so pick a darn good sounding driver if you want to make a line array (my advice).
_-_-bear