923 Haddonfield Road, Suite 300 Cherry Hill NJ 08002
+1 856-324-8258

Line Arrays Explained

Welcome to The Roadies of Color United (R.O.C.U.) Touring Professionals International Network Website.

Line Arrays Explained

By David Mellor

If you’ve been to a biggish gig or a festival in recent years, you’ve had the pleasure of hearing line arrays of loudspeakers in action. But why are line arrays the current ‘best practice’ in large–scale PA, how did they evolve, and will they ever filter down to more modest gig venues?

Here’s a chance to show off what you know about live sound engineering. Simply complete the following sentence: The function of a PA system is to…

That wasn’t hard, was it? But in case you’re struggling, the function of a PA system is to deliver your sound to the audience, and deliver it well. It’s as easy as that. But hang on, it doesn’t seem to be all that easy, does it? Whenever have you experienced perfect sound as an audience member? And when have you ever felt that your band’s sound has been delivered to the audience as well as it should have been? There must be additional criteria that need to be fulfilled to achieve satisfaction. And yes, there are. Three…

  • Adequate level, in relation to purpose (clearly, heavy rock music needs to be louder than a classical guitarist).
  • Low distortion, low noise and a flat frequency response.
  • Adequate clarity, in relation to purpose (speech requires near–100 percent intelligibility; all the words in a theatre musical must be easily understood; other forms of music may not need to be absolutely crystal clear).

Achieving adequate level is never a problem. It hasn’t been a problem since the 1970s, when PA systems as we know them today had fully matured. All you need is a recognition of how many watts you require for a particular venue, usually calculated by rule–of–thumb and reference to past experience, and the budget to hire enough amplifiers and loudspeakers. Achieving low distortion, low noise and a flat frequency response hasn’t quite been fully solved, although if the noise level of your PA is audible to the audience there’s a fault somewhere in the system: power amplifiers in general have a better signal–to–noise ratio than just about anything else you’ll find in the whole of sound engineering. The frequency response of PA loudspeakers, however, leaves a lot to be desired, and it is definitely true to say that the only thing that produces more distortion than a loudspeaker is the lead guitarist’s screaming Marshall on overdrive. But even though not all is yet perfect regarding the above points, most people find the sound quality of a decent PA system acceptable. And the typical sound of a PA has almost defined people’s expectations of what a PA should sound like. A circular argument, perhaps, but there’s a lot of truth in it.

There’s still one point left unanswered: that of clarity. It is possible for a PA system to be capable of detailed, analytical clarity within itself. But when deployed in a real–life concert scenario it sounds anything but clear. You must have experienced it yourself many times as an audience member — that fuzzy mush of sound that clogs up your ears, but you can’t really resolve it into music. Clarity, therefore, is the last unconquered frontier of PA. It is the last major problem that remains difficult to solve.

At this point I need to return to one of the requirements of PA that I previously said had been solved: that the PA system should be loud enough. There’s no difficulty in making it loud enough, providing you have the budget — but it has to be loud enough for all members of the audience, and that’s a problem that isn’t necessarily solved just by spending a lot of money.

There are two scenarios here: one where the audience are seated, the other where they are standing and free to move. If the audience are free to move, it is acceptable to have different levels in different parts of the venue. Those who like it loud will gravitate towards the loudspeakers. Those who perhaps want to chat during the show will move further away. However, if the audience is entirely seated it suddenly becomes much more difficult. You don’t want to deafen the front rows of the audience while leaving those at the back struggling to hear. If only certain members of the audience are delivered a level that is adequate, without being too quiet or too loud, the PA has not fully met its purpose. Let me therefore refine the requirements of PA into this simple statement: all of the audience should enjoy high–quality sound that is loud enough and clear enough.

Line Arrays explained

Cover The Audience, Not The Walls

A paramount rule of PA is to direct the sound towards the audience and not elsewhere. But how often do you see this rule flouted? The best and most classic example of this not being done was in several London Underground stations, some years ago. At the time, the tube network was decaying and falling into disrepair, so several stations were refurbished with bright, modern designs. Along with the visual aspects, these stations were given new sound systems too. Some bright spark designer decided that the loudspeakers should be mounted in cylinders (cylinder = tube, get it?) and several should be mounted at intervals along the platform, parallel to the platform and just above waiting passengers’ heads. The result was that from any point on the platform, you could hear every loudspeaker, with delays increasing with distance. It was, indeed, possible to stand as close to a speaker as you could and still not understand what was being said! This state of affairs wasn’t allowed to continue for long, and now the speakers point as they should — down at the passengers on the platform.

So the most important thing is to point the loudspeakers at the people in the most direct way possible. At the same time, consider how much sound is being ‘sprayed’ onto the walls and ceiling. The audience will absorb much of the sound energy that strikes them, meaning that it won’t be reflected to bounce around the auditorium and cause confusion. But the walls and ceiling are very likely to be reflective, so the more sound that goes in these directions, the more mush–inducing reflections will be created.

The EAW CLA37 column loudspeaker uses seven 3–inch drive units to achieve a coverage of 120 degrees horizontal x 30 degrees vertical, thus controlling the vertical dispersion tightly. It is suitable for speech reinforcement in large reverberant environments if several or many units are distributed amongst the listeners.

The EAW CLA37 column loudspeaker uses seven 3–inch drive units to achieve a coverage of 120 degrees horizontal x 30 degrees vertical, thus controlling the vertical dispersion tightly. It is suitable for speech reinforcement in large reverberant environments if several or many units are distributed amongst the listeners. In a situation where the information content of speech is of primary importance, the classic solution to intelligibility is to use many small loudspeakers and have them close to the people — obviously, pointing at them and not at reflective surfaces. This works extremely well and the information content gets through clearly. But this solution is not acceptable for a musical performance. The reason for this is that we expect a performance to take place on a stage. We watch the performers on the stage, and we expect the sound to come from the stage too. If the sound were coming from a small speaker mounted at just a couple of metres distance, up and to the side, that would cause a conflict between the visual and the auditory. Everything might be clear and intelligible, but we wouldn’t enjoy the performance.

So the multiple small speaker solution doesn’t work for performance. We need the sound to seem as much as possible as though it comes from the stage, and for this you can’t do better than actually having loudspeakers at the sides of the stage, like a great big stereo system. However, there are still potential problems…

The first problem has been mentioned already and has to do with directivity. Loudspeakers naturally have a characteristic directional response — almost omnidirectional at low frequencies, tightening to a focused beam at high frequencies. Put another way, anyone sitting directly in front of a loudspeaker will experience a reasonably flat frequency response, but people sitting further and further to the side will hear less and less high frequencies, so the sound will be increasingly dull. So the ‘big stereo system’ style of PA suffers in that it sprays the walls and ceiling with low–frequency and low–mid energy that reflects into a confusion of reverberation, and only select members of the audience receive sound with a good balance of frequencies.

A second problem stems from the lack of directional control. Because much of the sound is spread widely, beyond the width of the audience, energy is lost. The more sound spreads out, the more thinly its energy is spread, and therefore the more level is lost with distance. This is an important point. The reason a sound source becomes apparently quieter as it becomes more distant is primarily because its energy is spread out. Yes, some level is lost through absorption in the air, but not much. It’s distance that’s the killer. An audience member sitting a long way from the loudspeakers will experience a distant and therefore quiet sound, while audience members close to the speakers are getting their heads blasted off!

Let’s think in terms of light. Take a torch bulb. Intrinsically it emits light almost equally in all directions, so by itself it isn’t much use for finding your way in the dark. But put a reflector behind it and a lens in front of it, so that its energy is concentrated into a beam, and you will notice immediately that it is now usefully bright. You’ll also notice that the beam extends into the distance. So not only do you see the immediate area in front of your feet, but the area beyond where you direct the beam. The area of coverage is less, but you can now see where you’re going. If the same could be done with loudspeakers, there would be two benefits: one, that the sound is focused on the audience and away from reflecting surfaces; and two, that the sound retains its level as it travels. So the audience members at the back are served as well as those at the front, and the difference in level between front and back is much less.

Directivity Theory

If you understand the theory behind the directional characteristics of sound sources, you’ll be in a good position to understand PA loudspeakers and get the best out of them. There are two extremes of directionality, between which there are other interesting cases. One extreme is the point source, which is a source of sound that has zero size. OK, there’s no such thing as zero size, but in practice if a sound source is dimensionally smaller than the wavelength of sound it is emitting, it has the characteristics of a point source. The low–frequency output of a small loudspeaker would be a real–life example.

A point source emits sound equally in all directions. There you have it: all you need to know about the point source! Well, not quite all… but you’ll need a little imagination. Imagine this very small point source pulsating outwards momentarily, just once. A sphere of high pressure leaves its surface and radiates outwards, becoming larger and larger. The point source has put a certain amount of energy into this pulse, and that same amount of energy over time has to cover a larger and larger area, the surface area of that continuously expanding sphere. I could at this point bore you to tears with detailed calculations concerning the surface area of a sphere, energy density and stuff like that, but instead I will cut directly to the chase and say this: for a point source, sound pressure decreases by 6dB for every doubling of distance. We call this the inverse square law.

One mistake or over–simplification is that it is commonly said that all sound obeys the inverse square law. This is not so. Only sound from a point source obeys the inverse square law. Any sound source that is not omnidirectional does not obey the inverse square law. (If you get so far away from it that visually it recedes to a point, from your point of observation it will appear to obey the inverse square law, but in practical terms this is not relevant to PA).

A four–module array of Renkus–Heinz STLA/9 cabinets.

A four–module array of Renkus–Heinz STLA/9 cabinets.From this we can derive two interesting facts. The maximum rate at which sound level can decrease with distance is 6dB per doubling of distance. The only way sound can decay at a faster rate than that is if you actively do something to block it. Also, sound sources that are directional decay at a rate that is less than 6dB per doubling of distance.

It’s interesting to consider the opposite extreme. Would it be possible to have a sound source, the level from which does not decay at all with increasing distance? Amazingly, the answer is yes. It is possible to have a sound source that is so focused that it will cover an amazing distance with hardly any reduction in level. You want an example? I’ll give you two examples: an old–fashioned ship’s speaking tube, and a tin–can telephone. We call this kind of sound source a plane source. In both cases, the sound energy isn’t just focused, it is constrained to travel within an enclosed medium so that it cannot spread out at all. And since it cannot spread, no level is lost. (In practice, a little level is lost, but nothing’s perfect.) You can see that this is not a practical way of delivering your sound to the audience, so we will leave it as a curiosity, but a curiosity that demonstrates a useful principle.

The next type of sound source is the whole purpose of this article, and is the salvation of PA as we know it. We call this type of source — fanfare of trumpets — the line source. To understand it, lets go back to the point source for a moment. I said that the point source (which is omnidirectional) needs to be small in comparison with the wavelength of sound that it is emitting. The converse is true too: when a sound source is larger than the wavelength it is emitting, it becomes more directional. And the larger it is, the more tightly directional it is. So a really large sound source would be tightly directional. This is what we want: a source that can be focused and directed to cover the audience, but not wasted on other areas of the auditorium.

But imagine you’re a loudspeaker looking out from the stage to the audience. The audience in front of you are spread widely from left to right, but from top to bottom — in perspective, from the rear rows to the front — there is only a narrow spread. You can see the problem. If you made a large loudspeaker that focused the sound tightly enough to direct sound accurately in the vertical dimension, it wouldn’t cover the full width of the audience. And vice versa: if it covered the full width, you’d end up covering the ceiling as well, and we know that’s a bad thing.

The solution is to devise a loudspeaker that is tightly focused in the vertical dimension but spreads sound widely in the horizontal dimension. To do this, the speaker needs to be large vertically, but small horizontally. Like a column, in fact. And here we have it (bigger fanfare of trumpets): the column loudspeaker! Did I say ‘column loudspeaker’? Sorry, I must use the more up–to–date and exciting terminology: line array. They are both examples of the line source.

The Column Loudspeaker 

I often think that one of the best lessons of the past is not to go there again. However, the column loudspeaker has as important a place in the history of PA as the electric guitar does in rock music. Yes, really. One day there might be people who make a living as historians of PA, and they’ll be able to tell us exactly how the column loudspeaker came to be developed. Until then, my guess is that it developed by chance and was found to work effectively. It seems like a natural development for a 1960s band to have speakers at either side of the stage for the vocals. Then they decide they want to be louder and need speakers with multiple drive units. But speakers that are wider take up more stage area, so they choose speakers that are taller. The typical pub band of the 1960s would therefore have a pair of column loudspeakers, for vocals, that typically would contain four 10–inch or 12–inch drive units, sometimes topped off with a small horn (for example, the WEM Vendetta). Although they might seem primitive now, in fact they worked surprisingly well. The small horizontal dimension meant that the full width of the audience was covered, while the large vertical dimension ensured that the sound was ‘beamed’ to the back of the room. However, the next generation of bands working at a higher level of the business moved on to ‘bins and horns’. (A horn loudspeaker is the most efficient way of converting amplifier power to sound. A ‘bin’ is a bass loudspeaker, which is commonly in the design of a folded horn. ‘Bin and horn’ systems of adequate physical size can sound very good, but their directionality is not necessarily well controlled.) Small bands followed suit with similar but scaled–down systems, and the column loudspeaker was forgotten. Small column loudspeakers, however, continued very successfully in speech PA, such as for places of worship, where intelligibility is all–important (see the photo below). The ‘bin and horn’ system amounted to nothing more than the ‘big stereo’ commented on earlier, and directional control was lacking.

The next real development in PA technology was the centre cluster, much used in musical theatre. The centre cluster relies on another directional technology known as the constant directivity horn. The idea here is to combine multiple full–range loudspeakers, each of which is designed to have a consistent directional pattern over a wide range of frequencies. Horn loudspeakers can be designed to do this reasonably well. These full–range loudspeakers are arrayed together into a part of a sphere and mounted high up to cover the whole of the audience. Each member of the audience is delivered sound through only one full–range loudspeaker (apart, of course, from people sitting exactly on the dividing line between the coverage of two loudspeakers).

The centre cluster is outstanding for its intelligibility. It fulfils the criterion of directing sound only at the audience, and has the additional benefit that it forms a single sound source, therefore there is no possibility of hearing delayed sound from another loudspeaker somewhere else in the auditorium — at least, in a pure centre–cluster system. But there are two problems: the first is that ideally the centre cluster would be designed first, and then the auditorium designed around it! The second is that if each audience member is delivered sound (apart from the exception noted) by only one loudspeaker, plainly there is going to be a limit to how loud the sound can be. There will always be a role for centre clusters but, as we shall see, there are more flexible (literally) forms of loudspeaker distribution.

The Line Array

Although the column loudspeaker was effective in its context, it suffered from a lack of scale and a lack of science, each equally important. So to scale up a column loudspeaker to auditorium proportions took the best part of three decades. Still, we got there in the end. Here comes the science…

Going back to the point source, we find that level drops by 6dB for every doubling of distance. With the plane source, the level doesn’t drop at all. So is there an in–between condition where the sound level drops by, say, 3dB? Yes there is, and it is the line source, which in theory can produce a cylindrical wave, as opposed to the spherical wave of the point source. A genuine cylindrical wave will have 360–degree dispersion in the horizontal dimension and zero dispersion in the vertical dimension. Any real–life source is going to be an approximation of this, but if someone offered you approximately £100, you would accept £75, wouldn’t you?

A Meyer Sound 12–cabinet MICA array.

A Meyer Sound 12–cabinet MICA array.Earlier, I said that to achieve directionality a sound source needs to be larger than the wavelength it is producing. To achieve focus, or near–zero dispersion, which is a more stringent requirement, it needs to be somewhere approaching four times the wavelength. The wavelengths of audible sound extend all the way to 17 metres (20Hz) and beyond. But taking a reasonable lowish frequency of 170Hz with a wavelength of two metres (taking 340 metres per second as a nice round figure for the speed of sound), a line source eight metres high will be necessary. Quite tall! But at least we have a notion with some science behind it.

The next question is: how exactly do you make a loudspeaker that is several metres high? Currently, the way to do it is to stack multiple loudspeakers on top of each other. But instead of stacking 10–inch or 12–inch loudspeakers featuring identical drive units with poor HF response, as they did in the 1960s, each loudspeaker consists of LF and HF drive units and covers the full audio range (down to a reasonably low frequency). Also, rather than making one very tall cabinet, the modern line array consists of multiple small cabinets. The benefit of multiple cabinets is that you can assemble a line array that is as big or small as you like, or can fit in, or can budget for. You can also manipulate the shape of the array, which, as we shall see shortly, has significant benefits. Time for more science…

Since the line array is not actually one single tall–but–narrow drive unit, but is made up from discrete loudspeaker cabinets, one has to ask whether the individual units will couple together as though they were a genuine line source? The answer is yes, they will, but only where the drive units are separated by less than half a wavelength. This is easy for the lower frequencies, but more difficult to achieve as the wavelength shortens. As a benchmark, the wavelength at 400Hz is around 85 centimetres. So to couple at 400Hz the cabinets have to be less than 42.5 centimetres high. OK, that’s doable, but we are not even halfway up the audio band here.

Still, at least we know the criteria to aim for. The longer the array is, the more tightly directional it will be in the vertical dimension, and for individual cabinets to couple well into the array, they have to be small vertically. The better both of these criteria can be achieved, the more controllable the beam of sound from the array will be. A good point is made by Ralph Heinz of PA manufacturers Renkus–Heinz: “The answer to the question of whether a line array is a line source is ‘almost never’.” Heinz’s comment demonstrates that a theoretically perfect line source is virtually impossible to achieve. Only the best line arrays will come close.

Continue reading this article here:

Please follow and like us:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Follow by Email