Constant Beam Width Transducers line arrays


Paid Member
2004-10-07 6:05 am
If I were a guessing man (and I am) I'd guess that Geddes speakers would sound better , but Keele's more impressive. Geddes does use better drivers and a concept that is long proven.

As bear says, use darn good drivers. Audience does in their line array. Of course one advantage of all those drivers is that they don't have to work hard, each one so each driver is not so strained. I think that if Don Keele had move his crossover point up a notch, it would have been a lot cleaner. The heavy carpet was limiting the bass from the array.

But Earl has a point. Horizontal beamwidth was not constant from top to bottom of the array. It gets tighter toward the top. But within any given horizontal slice, it seems constant. I don't know of any speaker that has a constant horizontal beamwidth from floor to ceiling.
Hello from over the pond,

i am a little bit disapointed, because some of the last posts
read like a "collection of prejudices" to me.

Not referring to any particular i guess that

- small drivers = no good sound
- vertical directivity = nothing good for
- horizontal directivity not constant = concept does not work

is not really covering the issues.

If i may say something as a simple man with some
practical experience:

1) A multitude of small drivers may very well outperform a
single driver of same area. This is mainly because breakup
issues can be shifted towards higher frequencies.

Using drivers below the range where beaming occurs can
also have positive effects depending on the underlying concept.

If the problem with a multitude of small drivers is displacement
volume or fs too high for reproduction of deep bass without
excessive EQ:

Do what you would do with every speaker and use one or more

I do not think a subwoofer below 80Hz would spoil the concept
of Mr. Keele , would it ?

2) Since the horizontal plane is that one we are moving and turning
in normally as human beings, vertical directivity is a nice thing to
maintain a higher ratio of direct sound over larger distances.

Since room dimensions being smaller in height normally,
those are the reflections good to avoid first.

3) Horizontal directvity dictated from the driver size(s) and the baffle
is the usual approach with almost any speaker on this planet.
This may not be a good thing, but nevertheless it is inadequate to use
that argument to disregard the concept of Mr. Keele.

(It is the reason why i personally went for a dipole line array.)

Kind Regards
Last edited:
Line array no better than a single driver?? WOW- The motion of each driver in a line array is substantially less than a single driver at the same SPL- this guarantees lower every kind of distortion there is. So that is not better??? Then what is????

That is only true if comparing drivers of the same size - if you are comparing drivers of comparable total sensitivity and/or surface area that may or may not be true...

I only very very quickly browsed the 47 page paper, but it seems like the design is depending upon the floor "image" as being required for the array to function as expected?

Also, in a room with a low ceiling (how about any normal ceiling?) that the upper drivers will be reflecting?? Unless they are rolled off substantially, but that would have to be quite a lot.

As i said I did not read the 47 pages, but i have downloaded them, and will read in some detail later...

Anyone have any thoughts on this??

I only very very quickly browsed the 47 page paper, but it seems like the design is depending upon the floor "image" as being required for the array to function as expected?

At least the mirror image from bottom reflections must be
taken into account, they make up an integral part of the whole.

Buy one, get one free ...

Kind Regards
Also, in a room with a low ceiling (how about any normal ceiling?) that the upper drivers will be reflecting?? Unless they are rolled off substantially, but that would have to be quite a lot.

I would guess, tolerating some vertical dispersion and ceiling reflections
is the price to pay for "constant beamwidth".

And i think in this design price (tradeoffs) and what to get for it (benefits)
are calculated very sharply ...


Paid Member
2004-10-07 6:05 am
The basic concept does not use the floor reflection, it is twice the arc of what was shown in Ohio. It would be hung in free space. But Don realized that a floor standing speaker would effectively be doubled in length by the floor reflection. He placed a mirror on the floor to illustrate this. We saw twice the array.
In the hotel meeting room the carpet was thick with a heavy backing. That made for pretty good acoustics for everyone else, but diminished the array reflection. Don noted this several times, saying that a tile or concrete floor would work better.

The ends are shaded (attenuated). In the case of this array by -12dB relative to the center. The center of this array is at the floor (reflection extends it below the floor). So there did not seem to be much ceiling spray off the top drivers. The shading seems to pull the vertical beam in, that's one of the major features.
the practical speaker seems to depend on the floor reflection.

of course he shows quite nicely the effect of "floor bounce" on those mini monitors that are set up on stands! :D

One of the problems with this design is that it does rely upon the floor bounce (reflection/image) for performance. It looks like it needs to have a very hard flat surface to work properly at higher frequencies. Hard and flat because anything else will have slightly diffuse reflections and so not work close enough to theory.

So folks with a flat terrazzo floor are in luck! :D

I was most impressed with the "shaded array" that was merely vertical!

Someone earlier I think posted that the array was frequency tapered, but I read that it is not frequency tapered, only amplitude tapered. Interesting.

In practical terms I just do not see how this speaker is practical for home use, since it sends a substantial portion of its energy UP, and that will be reflected back - resulting in a messy looking picture (those neat simulations). Of course it might sound just great.

For me this was a very good paper to read, and as a result I see some definite opportunities to do things a bit differently than Keele elected to do (for home use)...

@ panomaniac: Are you sure that the array is attenuated by 12dB at the end, I seem to have read that it was less, but it was a fast read. Also the array is tapered only from the bottom to the top, since the image is the other half...

There is a whole lot of speculation going on in this thread about the CBT. I have read all of Don's papers, some multiple times, and I personally recommend that you start with the Ground Plane Paper. This will jump start you on understanding his approach because it starts with the full array and then shows how the floor image can be used for half the array.

I think it is important to point out that this technique can be used to design a speaker that can provide both vertical and horizontal control, but the design shown at the beginning of this thread is intended to have a wide horizontal pattern and controlled vertical pattern.

Above the lower cutoff frequency of the array, the vertical beamwidth of this design locks in and remains constant though out the audio band, the horizontal beam width is very wide at all frequencies.

One last comment before this thing times out, the top drivers are at -12dB, the shading occurs in steps of 0, -3, -6, -9, -12dB with a net loss of just under 3dB for the array.

The shaded array I have worked with uses synthetic aperture technology which results in half the energy radiated into the 20 degree tall vertical beam. The other half of the energy goes everywhere vertically (the other 160 degrees) and is not very important as the beam is 10dB louder than not in the beam. Horizontal control is left to the midrange and tweeter drivers designed to have the same frequency response on axis and 30 degrees off axis. These two together make a 60 degree wide 20 degree tall rectangular window for the listening area straight in front of the center of the array. As the mids are 6.5 inch and the tweets are 1 inch with lens the directivity of the speaker array is pretty constant from 200Hz-20kHz. First reflections in the room are completely unnoticeable because the direct sound is so much louder, at least 10dB in the main beam/lobe. This is done to eliminate room effects and items in the way of the sound, floor and ceiling bounce, and so on. The result is a speaker where at peak levels of 110dB at the listening position (10 feet) the midranges do not even appear to move. Image is stunning and clarity is extreme. Comparing the sound to highest quality headphones the headphones mask out a lot of the sound and loose a lot of the dynamics hiding popping "P" sound clearly audible on the array, as example.

There are a lot of ways to make arrays and of course some work better at some things than others and some solve problems that others do not, of course. It seems to me it is important to target statements about arrays to a particular style of array as the end results vary with particular designs so widely in this style of speaker system. I have never heard or tested anything so dynamic and alive sounding as a good array. All other speaker methods are much closer to dull lifeless and boring by comparison.


Paid Member
2004-10-07 6:05 am
All other speaker methods are much closer to dull lifeless and boring by comparison.

Yeah, kinda like real acoustic instruments. ;)

Just poking fun. But line arrays always sound "hyped-up" to me. A lot of fun and spectacular, but not realistic. Comes down to what you like - a matter of taste. Great for rock and other amplified music genres. It's a very popular style.
I have never heard or tested anything so dynamic and alive sounding as a good array.

At the conclusion of Don's presentation, he treated us to a recording of fireworks...the finale...the conclusion.

At the end of the day the CBT's were set up for one last go...radial aircraft engines...they were in the room with us, the air being chopped by the blades was palpable...and I have been around aircraft

Just like some food tastes good but is not healthy ? :D

But seriously, i cannot see line arrays lacking "reality" from
their working principle.

But i can see the common 2- or 3 way lacking reality by principle.

Like in everything only the real implementation can be judged and
there may be better and worse ones too.
Last edited:
My key point is there are a lot of solutions to playing music through speakers from a single full range, like a headphone, to these arrays. Every choice has certain advantages and disadvantages which "weigh in" on the final decision. I believe almost any design has merits not found elsewhere therefore no poopooing any one because some other does not do what that one does.

Everyone is right, it is a matter of taste. I happen to like my speakers to sound like live except in really good short reverb time room from where the conductor stands at the podium. Arrays give me that sound more than any other.


Your description of a shaded speaker is interesting, but not quite clear, perhaps you can expand on it and/or provide a diagram??

The idea that a 10dB down major lobe bouncing off a surface is not "noticeable" just isn't accurate. What you might say is that YOU did not find it objectionable??

Take the same speakers outside and listen - you might then just notice how MUCH of a factor the room and the stuff bouncing off the surfaces really is?


@ everyone

To keep this in context - I am an advocate of line arrays - having built many in the 70s... (again, see my website Archives section?).

I can tell you that there are many speakers that can "reproduce" very impressive sound levels with low distortion today, line arrays do not have a corner on that aspect.

That Don Keele's speakers work well, is fine with me.

Modeling of line arrays

Since there seems to be interest in line arrays, here is an AES paper I wrote on the topic. It is a little more basic/broad than Don's and might give you all some useful background.

Had to split the PDF into 4 parts.

David Smith


  • Line Arrays Part1.pdf
    193.8 KB · Views: 893
  • Line Arrays Part2.pdf
    195 KB · Views: 621
  • Line Arrays Part3.pdf
    185.3 KB · Views: 546
  • Line Arrays Part4.pdf
    158.6 KB · Views: 596
  • Like
Reactions: 1 users


Paid Member
2004-10-07 6:05 am
I believe almost any design has merits not found elsewhere therefore no poopooing any one because some other does not do what that one does.

Quite agree, quite agree. I have my favorites but am constantly surprised by others. I've gotten to the point that I just don't know what's going to come out of a speaker. There are some real surprises in store.

Now I have to put on my "grumpy uncle" hat and talk about the Keele array and demo. I was there, I heard it. He had done some of it the night before, too. It was big, it was loud, it was fun, but it was not all that great. I heard speakers, speakers that were straining at that SPL. And lots of overhang, if that's the term. I.E. cones continuing on when they should have stopped. At least 1/2 that was the sub.

I've heard those engine, cannon and fireworks demo things done sooooo much better. All the dynamics without the sloppiness. It's mind blowing when you hear it. To Keele's credit - his arrays did a great job for their small size. Bravo! <hat's off>

As Sum says, it's a matter of taste. Most speaker systems sound puny to me, and the big ones sound overblown. I hear a velvety, delicate smoothness and subtlety in live acoustic instruments and voice that I rarely hear on speakers. But for rock, pop, reggae, hip-hop who cares? Bigger is better. A matter of taste, for sure.