Often the theory worked, but the packaging was less than successful...

Often component loudspeakers are installed as though they were packaged systems, with the same physical configurations and misalignments that are inevitable in packaged products...

Back in the equivalent of the Late Cretaceous period of loudspeaker systems, say 25-30 years ago, the commercial loudspeaker market was made up of a relatively small collection of manufacturers. I was a contractor, I knew of most of the competing product equivalents in the various product catalogues. It was a situation that mirrored the Renaissance, where an educated man may have known all the facts that were known about the world. At that time it was possible for one contractor to know about all the loudspeakers that were on the market. Time passes, and we find ourselves here in 1998, at a trade show with a huge number of loudspeaker manufacturers offering products to the pro sound and commercial installation industries. What has changed in the last twenty or so years you may ask? That's what I was wondering, what has actually changed in the world of loudspeakers since the Big Bang of loudspeaker manufacturers in the 70's. Of equal interest to me was the question of what has changed in what we are expecting from loudspeakers in the past twenty plus years. And that raises the question of what we have learned, as an industry, about loudspeakers in the past twenty or thirty years, and how has that changed the devices on the market?

The period from 1972 to 1978 saw a huge proliferation of new and specialized manufacturers of loudspeaker systems, using components from the major manufacturers. These new companies were finding the first market niches for specialized systems in the fledgling pro sound markets for live music, clubs and touring. The major manufacturers began to respond with products of their own, based on the sound acoustical engineering principles needed to wring every possible dB out of moderate power handling drivers (this was the days of 100 watt continuous program power ratings for 15" bass drivers). Often the theory worked, but the packaging was less than successful, either from a roadability or ease of configuration point of view, and sometimes just because the products lacked cachet (no big name acts endorsed the product). The cottage loudspeaker industry, or more accurately, garage loudspeaker industries continued to tweak and experiment with all sorts of loudspeaker concoctions, using odd configurations of devices, components and boxes, and refining their niche positions.

In the 60's and early 70's, the majority of installed sound equipment had its roots in the cinema markets. The loudspeaker products were usually the very same products as used in behind-screen installations, put together in larger clusters to cover bigger venues than any of the original speaker system designers had ever envisioned. The expectations for SPL and bandwidth were also growing as the venues grew. The massive size and efficiency of the big multi-bass driver cinema horns, combined with multi-driver multicells, could provide movie audiences with adequate thrills and spills in a big theatre with only a 100 watt amplifier per channel. Those same speaker systems were being placed in huge indoor venues, and with the available power amps of the day, they were able to generate reasonable levels in the seats, but these systems generally wouldn't make your pant legs flap at 400 feet (an experience I had a few years ago at the introduction of the Community Leviathan II in an outdoor stadium). This was prior to the general availability of wide-band, high-power amplifiers, which became more accessible in the 70's. It would be safe to say that much of the installed commercial sound systems, right through the late 70's and early 1980's, owed a great deal to the heritage of behind-the-screen cinema loudspeakers. A wide variety of theatre loudspeaker components were used to assemble large loudspeaker clusters from. There had been developments of new horns to replace the aging multi-cell and radial/sectoral horn technology, and the low frequency packages became smaller as driver power handling and amplifier size increased. But the loudspeakers used in clusters would still have looked quite familiar to any service technician who worked on cinema loudspeakers.

Twenty years later, the commercial installation market is still seeing products that are spin-off's of another market. Instead of the cinema system market, it is the pro-sound and touring market that dictates the general style and approach of much of the loudspeaker product aimed for coverage of the commercial installation market. The bulk of the product development in pro sound and touring sound technology has been for lighter weight, higher power handling, less power compression, and now lower distortion. This has resulted in improved driver technology, power handling and reliability. The primary goal in the design of touring loudspeakers is maximum decibels of output per cubic foot of truck space, the widest bandwidth for the box count, adequate coverage adjustability and the necessary flying flexibility to handle the widest variety of venues without hauling additional gear in the truck.

Touring systems have to work in difficult environments; venues where they are often a hundred plus feet from the listeners, the reverb time may be over 5 seconds, and they have to overcome the background noise of a crowd of 10,000+ people having a fine time. Those are not the typical operating conditions for most installed commercial sound systems, and the design goals of most installed systems are quite different than a touring sound rig hung in a roadhouse or arena. Only a small percentage of commercial sound installations are of the large outdoor stadium or large indoor arena variety where the design goals of large concert type arrays may be applicable to the permanent install.

Packaged speaker systems rather than component cluster systems seem to dominate new pro sound and commercial installations, to the point where it can be difficult to find a contractor that knows how to install horns and bass systems properly. Often component loudspeakers are installed as though they were packaged systems, with the same physical configurations and misalignments that are inevitable in packaged products, a bit like packaged boxes that have become invisible. That kind of defeats the whole advantage of component arrays; the ability optimize the configuration of the devices in the array to provide the best alignment of devices while providing the required coverage and power density through the listening area.

Now before my e-mail inbox overflows, I'll say here that there are applications where packaged systems are useful, and applications where component arrays are the most useful. I will also say that there are manufacturers of packaged systems that take their development for installations more seriously than others, and put more effort into practical solutions. The problem is that the difference between the manufacturers that do and the ones that don't are not always obvious to all the purchasers. Marketing and price seem to win over technological suitability all too often.

When considering the use of full range packaged systems over component arrays, one will often find that the application of those two solutions are not truly complimentary. It is possible to substitute component clusters of equal driver quality for packaged speakers and achieve good results, it is seldom possible to substitute full range packaged systems for applications that are best suited to component arrays, and still get good results, there is not enough flexibility in driver location and position to reduce or eliminate problems in coverage and overlap. This really comes down to the difference in the design goal between most commercial installations and high level music reinforcement systems.

A properly configured horn and box "junkyard cluster" can sound just as kick-butt as a packaged box array...

There are many aspects of full-range packaged box cluster design that seem to be ignored...

One of the things that is very different about 1998 versus 1968 or even 1978 is the knowledge we have gained about the way loudspeakers behave in the time domain. Because we have been able to make time-domain field measurements of real speaker systems since the early 80's we know that signal misalignments in loudspeaker clusters create all sorts of horrific problems in coverage and response uniformity. As an industry, we found that "junkyard clusters" with lots of horns and boxes often sounded bad because of the excess overlap of devices, misalignment between adjacent devices, and the generally chaotic component organization that often had to happen to fit the support structure. The resulting coverage and frequency response of a cluster of this type varied widely with seating position. There was a period of time when clusters of this type were being tweaked and rearranged to provide better alignment, and reduce the overlap, etc. I remember being involved in one such cluster redesign that resulted in removing a dozen or so VHF tweeters plus four or five mid horns and actually having better bandwidth and uniformity with the same SPL in the seats. TEF measurements showed the before and after results, and the listening tests confirmed what the measurements showed. A properly configured horn and box "junkyard cluster" can sound just as kick-butt as a packaged box array if the component selection is of equal quality, and the power allotment is the same.

In the 90's these same tweaked component clusters are often old enough that they are being removed and replaced with full-range packaged speaker system clusters where many of the variations in coverage and response are being re-introduced because of the spacing of devices that occurs with the hanging hardware in place, especially where the vertical coverage requirement is larger than a single box will cover. I am guessing this is some sort of nostalgia for the 60's and 70's, since it seems to ignore what we have learned about the behaviour of multi-driver speaker clusters in the past 15 or 20 years. The more things change the more they stay the same. In the old days clusters were installed with those problems because we didn't know better, now they are installed because they are easier to install, and have sexier marketing. This isn't just my opinion, this is often what we're finding when we are asked to measure new systems that are being installed using full-range packaged boxes, after the client finds that speech intelligibility isn't as good as they'd like it to be and the speech quality varies greatly in every seat.

Many of the traditional manufacturers of loudspeaker components for clusters have also brought packaged systems to the contractor market because a sizable chunk of the market dollar is going that way. These systems are often of the generic trapezoidal shape implying array-ability, and yet the coverage angles of the loudspeaker components would not allow them to be arrayed in the most physically obvious fashion without excessive overlap of coverage. This tends to lead to the boxes being misapplied by contractors or designers who may not understand how loudspeakers work together when arrayed in close proximity. Despite the widespread availability of software to assist in predicting the performance of loudspeakers, we still see many design/build projects designed from some empirical method, without due consideration for loudspeaker interaction. Even the most rudimentary modelling software would show the problems that can be expected from some of these designs.

Both large and small manufacturers are producing these types of boxes. The lower cost "contractor" versions of the packaged multi-way full-range systems sometimes inherit the physical appearance of the touring systems, but not the same driver compliment, or electronic processors used in their full-size, full-cost cousins to provide signal alignment and equalization. It's true that most contractor boxes don't need the 600+watt woofers, they will likely only see 50-100watts of voice and music program, so that, in itself, is not the problem. It's the driver configurations that are often the source of the probelm. Sometimes these boxes have simple passive crossovers built in to keep them within a certain price point, and have no correction for signal offset between low and high frequency devices, generating new problems in the crossover region. Despite their touring box appearance these products are often intended as stand-alone, single box, wide-angle coverage solutions, and are not intended to be arrayed at all. These are often marketed as universal solutions without adequate training from the manufacturers in their correct applications.

There are many aspects of full-range packaged box cluster design that seem to be ignored, the primary ones being loudspeaker behaviour off-axis, and the coverage below the high frequency horn cutoff. When I have seen computer modelling coverage plots submitted with design/build proposals, they tend to be plotted at 2 and 4kHz, where the horns have the best directivity. The plots often don't account for the loss of directivity below 1kHz and the narrowing at 8kHz and up. The LF driver spacing usually plays havoc with the coverage between 250Hz and 1kHz in multi-box clusters. Many of the cluster designs show a real lack of understanding of what happens when the sound leaves the speaker system.

Where the listening environment is adequately reverberant, the holes in the coverage do not generally present an audible problem. Very large clusters with wide coverage, operated in highly reverberant arenas, will tend to place most listeners well beyond critical distance, in the reverberant field where variations in the direct sound are much less identifiable. In rooms with an occupied reverberation time below 1.5 seconds, the variations in the coverage and response are plainly audible where they occur. So the typical contractor market venues where the small full-range packaged boxes are most commonly used; churches and drama theatres, often exhibit the most audible problems. For music playback this is not usually a big problem, but it can be a killer for intelligible speech and voice quality.

In the late 70's we began to discover what it took to make loudspeaker clusters work. In the 80's, when we had the ability to see what was happening in the time domain, we changed the way we put clusters together. Some major manufacturers that produced speakers for the contracting market took the contracting market seriously enough to offer training courses in system design, especially loudspeaker cluster design. That is a trend that has pretty much disappeared, and to the detriment of industry. We have more time domain based measurement systems available than ever before, and there are still very few (by percentage) contractors who are making use of those systems to optimize their loudspeaker system designs. That knowledge and understanding of proper loudspeaker cluster design that was hard won in the 80's has been lost as many of the contractors and designers now in the industry have come from other segments of the market where the issue was never important. In that respect convergence is setting us back 10-15 years in loudspeaker system design. What the contractor loudspeaker market needs now is more training for the people that make use of the products so that the designers and contractors will use it properly and effectively.