Schools Of Thought
Belief System #1: Many sound system designers and audio practitioners have adopted the notion that if two trapezoidal loudspeaker enclosures with a stated angle of dispersion – let’s say 30 degrees for this example – are placed adjacent to one another at a 30-degree included splay angle (this means that the side walls of the trapezoidal enclosures are probably 15 degrees each), then the result will be a seamless 60 degree system. Or, if three enclosures are used adjacent to each other, they will form a seamless 90 degree system, and so on.
If this was as complicated as it gets, then all 30-degree loudspeakers would be the same as all others. Moreover, the total power output of the array could never be greater than that of each array module. But that’s absolutely not the case.
There is no such thing, in practical usage, as a 30-degree loudspeaker, nor a 45, a 60, or a 120 (though there are plenty of 360-degree loudspeakers; most are known as subwoofers). To understand why this is true, let’s look at how directivity is stated in loudspeaker product literature.
The angle of dispersion that defines acoustical coverage is almost always based on -6 dB “down” points in which the off-axis energy is -6 dB (less) than the on-axis energy. And that’s usually measured at 1 kHz, or at some frequency in which the HF horn (and MF horn, if applicable) exhibit some measure of control. But this nearly ancient way of characterizing a loudspeaker’s dispersion ignores the vast differences between LF, MF, and HF. Acoustical behavior is not cut and dried.
At lower frequencies, the LF content will get wider and wider in dispersion until it becomes essentially omnidirectional, under all but the most abnormal of conditions (think 60-foot long horns…abnormal, right?). Very long wavelengths (low “E” on a bass guitar is about 30 feet in length) need either very large horns or very large arrays to provide pattern control. Think of it like this: doing much of anything that might control LF dispersion takes a lot of building materials.
Coverage provided by vertical and horizontal L-Acoustics ARCS arrays.
So here we have our first conundrum: the directivity pattern of a loudspeaker, thought by many to dictate what an array will (or will not) accomplish, typically varies by about two orders of magnitude from a loudspeaker’s LF range (assuming it can reproduce 40 Hz) to its HF range (assuming it can reproduce at least 4 kHz).
And that brings us to our second conundrum: an important property of dispersion, and especially how it affects arrayability of like devices, is often overlooked. The property is what happens after the -6 dB down point. A given horn might cut off drastically in amplitude, whereas another might fall off slowly and gently as the dispersion angle increases.
In other words, a horn stated to be 90 degrees at 1 kHz might provide very usable response at 120 degrees, perhaps being only -9 dB than its on-axis reference amplitude. And that could be a good thing if the response remains reasonably flat. The off-axis seating is likely to be physically located much closer to the loudspeaker than the on-axis seating, which means that additional “fringe-fill” loudspeakers may not be needed.
By contrast, another 90-degree horn or horn array might be as much as -24 dB lower at 120 degrees, making its broad-band off-axis response rather unusable. But more to the point, these widely varying horn characteristics will not behave the same when configured into an array. And this leads us to…