Everything you ever wanted to know about Line Arrays ( but we’re afraid to ask)

OK, let's unpack line arrays. The history of line arrays in domestic audio goes back some time. Once upon a time, when amplification was expensive, people were looking for ways to make louder speakers that were more efficient. Using lots of drivers is an excellent way to accomplish this task, and the most cognitively simple way of making a big array is to arrange a bunch of drivers vertically. That's what you see in the pipe-dreams speakers and many others dating back to the 60s or before. At some point it became accepted audiophile wisdom that such an array was 'cylindrical' if it was sufficiently tall in the room, and the fall-off of volume vs. distance would be different from a normal speaker. In a way, it would not abide by the inverse square law. As always, the reality is much more complex, but back then nobody was doing 3d speaker measurements, so their illusions persisted through the decades.

The problems with line arrays? Well, they work okay for low frequencies, at least as well as any other speaker below the transition region of the room, but at high frequencies you get a rather chaotic interaction between all the drivers. Ideally, you want center to center spacing to be very low, but in reality the path length difference between your ear and the middle / end tweeters is always going to be significant. As a result, not a very hifi solution, but if the power response, ie the total radiated energy, is where it needs to be, they can sound good, and you can run them with a tiny amp.

So what do we see with PA systems? Well, don't quote me on the history, but if I remember correctly, Don Keele was doing some research into either microphones or speakers developed by the navy. Lots of acoustics doesn't actually deal with air; a lot of stuff is discovered in water. Thank the Russians I guess. Anyway, they developed a way of arraying these transducers so they either sensed or produced a very coherent wave, without the interference problems intrinsic to a bunch of drivers on a plane.

Keele took this idea, which used a spherical surface, and cut it down into a single curved array. This curved array was 'shaded' meaning the drivers at the extremes were quieter than the ones in the middle (maybe I have it backwards.) This is pretty clever; you can get a coherent line array, promising everything audiophiles promised back in the day, simply by curving the line array and making some of the drivers quieter. There are still problems with center to center spacing, but if you only want the speaker to play above 200hz or so, you can use a ton of tiny midrange drivers, and maybe even skip the tweeters. Parts express sold a kit which used this concept, and the room measurements of these speakers are remarkable and probably better suited to a large home theater than any other design. Don Keele was an engineer at EV but I'm not sure who he did this research for, needless to say, the cat is out of the bag, and these systems are ubiquitous now.

What you see in professional sound are arrays which use shading and also delay. If you delay the speaker, you don't need to curve it, or you can curve it differently. I'm not an expert on these systems, but they essentially accomplish the goal of high efficiency (especially at high frequencies - remember it's difficult to combine HF sources, so you either need a single HF source with incredible power handling, or a very clever method of combining HF sources). These systems generally have a HF transducer in the middle and two mids on either side. I think people like them because you can make the system bigger or smaller depending on the need for output and coverage. I personally think these systems sound terrible but that's just me.

So where does that leave us? Line arrays, like every other esoteric speaker design, tried to address a real design problem while sacrificing a bunch of other stuff. In this case, they sacrificed nice treble dispersion and price point for high efficiency and...not much else? Fidelity is not great compared to single radiators, but if you want super even coverage for thousands of people, it's the best technology out there if you combine it with some snazzy cardioid subs.