Why Large Sound Systems Need to Go Vertical

By Marty McCann

In this article I will address the issue of vertically arrayed sound systems, explaining why arraying the larger sound system vertically can immensely improve upon the performance of the same number of components spread out in the horizontal plane. I have been both a student and practitioner of concert-level sound-reinforcement applications for more than twenty years now. I have closely observed major-sound company systems and conducted numerous experiments of my own. One problem, as I see it, is that many smaller sound companies don't array their loudspeakers in a tall-enough vertical array. They tend to spread the system out in the horizontal plane, which increases the overlap of the constant-directivity horns and results in less-than-ideal polar patterns.

The best system (by another manufacturer) for large venues that I have ever heard is the ShowCo Prism system, which is arrayed four or five times taller than it is wide. Granted, it is more work to stack the loudspeaker system in the vertical plane. In most cases you must encase the speaker array in scaffolding, and/or band it together with nylon webbing to insure that the system will not tumble down on the audience. Some of the larger touring sound companies have a pneumatic loudspeaker tower system that enables them to attach each row of speakers beneath each other while they lift the entire array. The ShowCo Prism system consists of several rows of splayed (arrayed) loudspeakers attached to a circular steel tower.

Today, just about every major touring sound company employs its loudspeakers in an array that is at least four to five times taller than it is wide. This is the optimal method of providing large-scale sound. When the system is arrayed vertically, it becomes one large line radiator that does a much better job of projecting a correlated sound field to the farthest seats. When the loudspeaker array is raised vertically, as opposed to horizontally, you can obtain the proper coverage by providing individual far, mid, and near-field segments of the array. I have seen many small companies attempt to supply outdoor-festival sound over the years, and typically their loudspeaker stack is four or five times wider than it is tall. This is not the best way to get the most performance with a given set of loudspeakers.

To array a system vertically, it requires considerable scaffolding to surround and secure the stability of the vertical array. ShowCo solved this problem by engineering a safe and secure means of quickly hoisting its vertical array based on a custom designed pneumatic lifting tower, which doesn't require any scaffolding. The top of the tower consists of a large "wagon wheel"?looking circular metal frame, which is attached to a center metal pole that is telescoped (raised). The loudspeakers are rigged in layers to the bottom of one another as the tower is raised. Today, most of us use constant-directivity high-frequency horns in our touring systems, which are unsurpassed for propagating high frequencies. Before the introduction of constant directivity (CD) horns in the late seventies, every exponential horn had the same problem, i.e., the higher frequencies in about the last two octaves came exclusively out of the throat of the exponential high-frequency horns. As a result, the high frequency information was very concentrated, or "beamy," i.e., the higher frequencies beamed like light from a flashlight and could not cover as wide a horizontal angle as the lower frequencies.

One mistake many sound engineers make, with systems large and small, is allowing their constant-directivity high-frequency horns to overlap too much in the horizontal plane. The medium of sound is the air itself. Sound propagation actually entails increases and decreases in the barometric pressure. When the loudspeaker moves forward, it increases the air pressure in front of the speaker. When the speaker moves backwards, it acts a bit like a vacuum pump, causing the air in front of the loudspeaker to become rarefied, or less dense. Each audio cycle that can be reproduced consists of a compression of the air molecules followed by a rarefaction of the air. Imagine two billiard balls that are rolling down a pool table at an angle to one another. Let's imagine that they are going to collide and that their paths form an angle of forty-five degrees. Once they collide, they will move in two different directions, forming another forty-five-degree angle away from the point of collision. In acoustics, we say that the angle of incidence (arrival) is equal to the angle of coincidence (departure). If two billiard balls traveling toward each other collide and move apart, forming the same angle as when they hit together, what makes anyone think that colliding air molecules would perform any differently?

Another important factor often overlooked by the less-informed soundperson is the sensitivity rating of the components within a multi-way sound system. With more efficient loudspeaker enclosures, fewer speakers and less power are required. The Peavey HDH(TM) series enclosures are still the best tools that we offer for large-scale sound reinforcement. Just because an enclosure is horn-loaded, it does not necessarily mean that it is high in efficiency. Let's use the mid-frequency horn of the HDH-1 or HDH-4 as an example: with one watt of power applied, the MB-3 horn in the HDH product will produce 109 dB of SPL at 1 meter away from the enclosure. We and other manufacturers also have mid-frequency horns that produce 101 dB of SPL (1 watt @ 1 meter). The MB-3 horn is rated power-wise at 700 watts of program material. At 700 watts of power the MB-3 can produce 137 dB of SPL (10 log 700/1). A mid-frequency enclosure that exhibited a 1 watt @ 1 meter sensitivity rating of 101 dB would require 4,500 watts of power to produce 137 dB of SPL (10 log 4,500/700 = 8 dB) to make up for the decibel difference in sensitivity. I don't know of any single loudspeaker enclosure with a maximum program rating of 4,500 watts.

Now that I have provided this preliminary information, let me show you an example of a Peavey large-stadium sound-reinforcement system. I went to Dakar, Senegal a couple of years back and spent a week in a 100,000-seat soccer stadium, working with the components of a large Peavey HDH sound system. We set this system up employing several different configurations of the individual components. We started out with the system configured in a typical horizontally spread arrangement. Actually, the customer's own sound engineers had already configured and used the system this way before my arrival. We set the left side of the stadium up in the manner outlined in the following section and compared the two different arrangements, i.e., we conducted an "A/B" listening test. I had drawn up six different approaches for the number of enclosures that we had to work with. We then proceeded to try each of the other five versions of a vertical array on the right side and compared it to the left side. Some of the versions sounded better than others, but none of them could match the performance of the first vertical array on the left side.

Although the HDH enclosures are trapezoidal, when you pack them tightly with the horns side by side, the horns still overlap considerably, which causes cancellations and interference. In the system that I devised, I placed an HDH-3 enclosure between each HDH-4, which resulted in a nearly ideal arraying of the system. The real key to the performance of this system was the multiple PC4-XL digital electronic crossovers I used to calibrate the system. In this case, I employed two per array to calibrate the near- and far-field components. Since then, I have conducted further experiments with multiple PC4-XLs on smaller systems and have come up with some very interesting applications, where I would use one PC4-XL for each bandpass.

The vertically arranged components easily out-performed the same number of components spread out horizontally. The point of critical distance, where the direct field and the reverberant field become equal in level, was extended far beyond my predictions. I thought we would get at least a 50% increase in critical distance. Well, we increased the point of critical distance by a factor greater than 200%. Of course, the vertically arrayed system had been aligned using the digital-delay capabilities of the PC4-XL. We experimented with other versions of vertical arrays, but the one outlined here was the most efficient.

I would like to thank Ibrahima Ndiaye, the owner of Studio Demille in Yoff, Senegal, for giving me the opportunity to configure his Peavey stadium-sound reinforcement system. After returning from Africa I was contacted by our distributor in Barbados, A&B Musical Supplies. Norman Barrows, the owner, was thinking of updating his concert-level sound system. I gave him the proposal and told him to look it over and get with me if he had any questions. Well, he liked what he saw and ordered the exact system. When he took delivery of the system, he arranged for a conference call between his sound engineer and me. I had his engineer in Barbados sit with the PC4-XL digital crossover in his lap while we spoke on the phone. I also had a PC4-XL in my lap. In the course of about a ninety-minute phone call, I told him everything I knew about the digital crossover and helped him set it up for their system.

In December, I did a presentation on the digital crossover at the week-long Peavey Advanced Sound Reinforcement dealer seminar. It took more than three hours to cover the new PC4-XLA. Norman has been using this system for a couple of years now. He enhanced the original system by doubling up on the UDH(TM) Subwoofers and adding two more vertical rows to expand the horizontal coverage.

That is the story behind the R&D involving the system that I am now going to outline. This proposal includes the components that I would use if I were building the system today. The drawings of the racks and the signal flow of the crossovers are from the original system used by both Ibrahima Ndiaye and Norman Barrows.

Click for larger image (46k) Click for larger image (45k)