Study Hall

Supported By

Loudspeaker Advancement: Part 1 In A Series On The Evolution Of Sound System Optimization

A walk through key optimization developments over the past 40-plus years, starting with loudspeaker technology.

We live in the present and plan for the future. Every once in a while it can be interesting to look back at our history and see he we got here. This puts things in perspective and helps us see what’s coming.

The essential challenges to tuning a sound reinforcement system, a process we now call optimization, haven’t changed in 40 years, and neither have the laws of physics.

But the tools and techniques of the trade have changed dramatically, albeit incrementally, over time. As a person who has been in professional audio for more than 40 years, here’s my perspective of how it was and the journey to the present.

Let’s take a walk to front of house and meet our modern sound system. We see an engineered loudspeaker system and a multichannel digital signal processor loaded with all types of filters, delays, and more (maybe too much more). There’s an acoustical analyzer ready to guide the optimization process and an experienced operator with a step-by-step plan for system tuning.

Not too long in the past, none of these things would be found there. Just 35 years ago we would likely find a collection of various custom loudspeaker cabinets built in the rental company’s shop. The extent of the “signal processing” would be an analog crossover with fixed filter slopes and a graphic equalizer.

At best we would find a primitive analyzer that no one trusted and little, if any, methodology or scientific process. All these aspects of our sound system have evolved in the period since, each in their own way and together as a whole.

I’ve been an eyewitness to the evolution of these tools and techniques into their present mainstream forms, as were countless other engineers. We can explore this history together by following five primary threads: the evolution of loudspeaker systems, signal processing, acoustical analyzers, analysis systems, and the methodology of optimization.

This is a personal perspective, not an authoritative history (which would be an entire book in itself). Every veteran of the pro audio industry has played some part in our collective progress, for which I am grateful. I encourage any and all to share their journeys with us and enrich our knowledge and respect for our past.

The central thread in my story is a 32-year relationship with Meyer Sound’s Source Independent Measurement (SIM). This was the first analysis system capable of measuring the acoustic response of sound systems during a concert using the music as the test signal. The analysis system approach originated in SIM has helped lead to the analysis programs that are now so common that most live sound systems use some version of it today.

Stone Knives & Bear Skins

It may be difficult for the modern engineer to visualize the crude tools of the 1970s. The only commercially available professional loudspeaker systems were targeted for fixed installations such as cinemas and “public address” systems, and weren’t the least bit ready for trucking, stacking or flying. The enclosures were not ruggedized with internal (or external) steel framing or integral rigging.

An “old school” column loudspeaker, the Shure Vocalmaster, which was introduced in 1967 as the first complete system (mixer, amplifier and loudspeaker). Legend has it that the Beatles used this at Shea Stadium concert (1965). It was not, but the system on the field was not much more powerful than the Vocalmaster shown here. We’re using it for our “Let it Be” concert on the roof of a dorm at Indiana University in 1975.

Popular music concerts needed portable power on a scale beyond the imagination of the loudspeaker manufacturers. Rental companies innovated in developing mobile touring boxes while several manufacturers transitioned into the role of supplying the driver components inside them.

At the time I entered the pro touring market, systems were described by which rental house made the boxes and the maker of the drivers, i.e., “a Showco system with JBLs.” Nobody (at least that I knew of) was touring arenas with off-the-shelf boxes from a manufacturer. That was for high schools, cinemas, and “Hair” on Broadway.

Loudspeakers did not fly. They came out of the truck and stacked on the stage. The only ones that did fly were designed by consulting firms, usually comprised of a big pile of horns and a few woofers that we termed “flying junkyards.” You can still see a few of these gathering dust in the ceilings of old arenas.

It got loud in front of our ground-stacked systems. It wasn’t “hi-fi,” it was rock ‘n’ roll. It was supposed to be rough and edgy. We all knew that you could either have high power or high fidelity but not both. The tradeoff in favor of power was simply accepted as normal as long as it got loud. Really loud.

The arena rigs I toured with in the 70s while at Showco and FM Productions were 4-way systems with 15-inch woofers, 12-inch woofers, horns and tweeters all in separate boxes. We ground stacked them on the stage and pointed them in the general direction of the audience.

For bigger shows, the stacks were taller and wider. Frankly, one of the key sound design principals was making sure it didn’t fall over. Other companies, most notably Clair Brothers, made single boxes that were complete 4-way systems, but again, the same quantity and stacking principles applied.

I remember the first time we flew an arena system. We stacked the loudspeakers into a big steel basket with a plywood floor and up it went. For real. More “sound system in a freight elevator” than “flying system” but the progress toward level uniformity was amazing. It did not have to be insanely loud in front to be stupid loud in back!

A 1979 concert with FM Productions at the Greek Theater in Berkeley. We are struggling with a custom 3-way box that has an HF horn tacked on the top of it (a primitive attempt at time-alignment). The box had minimal handles cut into it and an “L-track” screwed onto the side for rigging. (Credit: Clayton Call)

The signal processing of the time was comprised of a 4-way crossover that drove the whole system, and simple limiters/compressors. Phase alignment was mostly just talk, since we had no tools to control it (no delay lines) or quantify it (no analyzer). This left us with the visual version: physically lining up the boxes as best we could. Again, the “don’t let the speakers fall over” principle came first.

Level adjustments were done at the crossover or at the amplifiers, the latter being a notoriously difficult method to obtain consistent results. Typical level settings for power amplifiers in that era were “2 clicks down” or “3 o’clock.” We had very little idea what the actual gain values were or how two different models related to each other. (By contrast, the modern amplifier can be software controlled in precise settings read in dB.)

Jorma Kaukonen at the 1980 Berkeley Street Fair. This era marks the beginning of the transition from custom and user settable loudspeakers to systems (processed). I was mixing monitors and had the chance to compare the new “processed” system, the Meyer Sound Ultramonitor (on the left) to a pair of conventional “do it yourself” loudspeakers. The single UM-1 was much clearer and more powerful than the old school pair. There was no going back. (Credit: Clayton Call)

Making Progress

The concept of the engineered system barely existed in the 1970s. Custom was king and users were expected to be able to add their preferred touches to the settings of the crossover and other electronics. Engineered systems, by contrast, fixed these parameters to precise settings selected for the particular combination of drivers, horns and enclosures.

This doesn’t sound radical now, but it sure was then. My first encounter with an engineered system was the rental company McCune Sound’s JM-10, designed by a young John Meyer. I stood there stunned as one of my long-held beliefs was shattered: yes, you could have high power and high fidelity at the same time in an arena.

The 1980s were a time of tremendous evolution in loudspeaker technology. For one thing, the dinosaur boxes with separate lows, mids, horns and tweeters gradually became extinct. Basically the industry settled toward two box types: “full range,” i.e. 2-way or 3-way boxes that covered from 70 Hz on up, and subwoofers for 100 Hz on down.

Manufacturers also began to make ruggedized enclosures ready for touring. The engineered systems up to this time had been kept as exclusive and proprietary by the rental companies. We now began to see this approach become non-exclusive as the manufacturers embraced it.

The integration of multiple components into a single box with fixed physical characteristics set the stage for the next level: the introduction of “processed” systems. Loudspeakers were sold as a “system” with dedicated loudspeaker controllers that provided fixed crossover settings, amplitude and phase response correction, and limiters. All of these parameters were pre-optimized for the specific driver/box combination.

It was the beginning of the “plug and play” paradigm that has become the industry standard today. It may be surprising to know that this was highly controversial at the time. Many engineers felt that the manufacturers were limiting their options for optimization by locking out parameters such as crossover frequency and slope, etc.

The main system shown here for the Grateful Dead at the Greek Theater in Berkeley (1984) had minimal subdivision, entirely driven by a single channel of equalization with no delays or relative level adjustment. The standard optimization options of a modern system (macro and micro delay, relative gain adjustment and separate EQ for the uppers/lowers/side fills/etc. were not available options at that time. (Credit: Clayton Call)

Part of the resistance was fear of the unknown, since what was going on inside the proprietary controllers was mysterious. Another part was the feeling that manufacturers were taking away a part of their tool set. Many engineers took pride in their personal crossover setting skills and did not want to surrender that control to others.

It took a long time for them to realize that a complete box with fixed, known parameters is better engineered in a manufacturer’s research lab than on the job site. The benefits of fully engineered systems over the custom recipe from Joe’s Garage became so apparent that the custom boxes of the Wild, Wild West era rode off into the sunset.

The processing and amplifiers for the system at Orchard Hall in Tokyo, Japan in 1988 or so. This system had an abundance of subdivision: left/right, upper/lower, inner/outer, and so on. The single-channel loudspeaker processors are in the rack on the left while the amplifiers are in the rack on the right.

Bringing It Together

Modern professional systems all have fixed pre-optimization. We buy “systems,” not just loudspeakers, and expect them to be fully engineered with optimized crossovers, pre-aligned frequency response, appropriately scaled power amplifiers, and dynamic protection.

They come in two varieties: self-powered and those with external amplifiers containing dedicated presets. End users expect to deliver a line level (or digital) signal to the system that in turn results in predictable output.

Each level of system evolution enables an evolution in optimization. How could we optimize a crossover in the old approach where horns and woofers were stacked up next to each other? Even with the best analyzer and digital processing there’s no sensible solution to that challenge (except a dumpster).

The standardized linear loudspeaker system opens the door to optimization. The sensitive crossover points have a fixed solution, which creates a known element that can be utilized to build arrays.

We can learn lessons about the coverage pattern, aiming, splay angles, compatibility with other models and more. Results can be predicted in advance because the response of the loudspeaker is standardized.

It’s personally embarrassing to think of how little I knew about loudspeaker array behavior back in those early days, but then again, it’s a subject complicated enough for an entire book. Even a simple, small array of identical elements is very complex, but the outcomes are predictable.

Known elements with known spacing and angular orientation will yield predictable results. This holds true with standardized engineered loudspeaker systems, but good luck with a system built in your garage and/or with wild parameters such as unmatched loudspeakers and processing.

A highly subdivided center cluster in a symphony hall with extremely complex 360-degree seating (and only one hang point). The front mains have upper/middle/lower subsystems with inner/outer. There are three subsystems for the sides and another for the rear. A total of nine channels of signal processing (EQ, level and delay) were needed. Each cabinet has independent pan and tilt to help facilitate optimal vertical and horizontal aim and splay.

Evolved loudspeaker systems exist at all power scales, running from Bambi to Godzilla and everything between. We see similar coverage patterns from 2-way systems whether they incorporate 15-inch or 5-inch drivers. They cover nearly the same frequency range, with the “big boys” reaching maybe an octave lower (9 versus 8 octaves).

The difference in power scale is gigantic though, which allows us to use a size proportional approach to achieve similar results. Recall the old school design principle: bigger venues use bigger stacks of the same stuff. The new paradigm is proportional scale: the same quantity of boxes can be used for a small or large venue, but we proportionally scale the elements up in size/power.

We define loudspeaker systems as coverage shapes and power scales. We can get 80-degree by 50-degree dispersion in all sizes, which makes it a scalable building block for both coupled and uncoupled arrays. After all, one person’s main is another person’s front fill. The modern line array is engineered to couple in large quantities in the vertical plane, but again the scalar paradigm applies.

This provides a picture of loudspeaker system evolution. Next time I’ll focus on the signal processing that drives them.

A modern era line array system (and the author) at the Appel Room, Jazz at Lincoln Center in New York. The main system is subdivided into four channels of signal processing (sub +3), and the various fill systems have independent processing as well.
Study Hall Top Stories