When expounding on the technical subject of sound reinforcement, there are plenty of great minds who can argue the merits of FIR versus IIR filters or explain the use of partial differential calculus in solving the wave equation. My job is to bring all of this down to earth.
With that goal in mind, I want to discuss a simple way to think about sound reinforcement systems. I will argue that the often complex and lengthy process we go through to design, install and commission a sound reinforcement system can be reduced to four questions and their answers.
This simplification has its limitations. We're not going to magically eliminate our need for tools like EASE or TEF. We're not going to put any acoustical consultants out of business. But, starting from a foundation of the four questions, we should be able to find our way through the process of system design and evaluation with a greater sense of confidence. Perhaps more important, the four questions may help our customers gain at least a basic understanding of what we do.
The Four Questions
Question 1: Is it loud enough (and not too loud)?
Question 2: Can everybody hear?
Question 3: Can everybody understand?
Question 4: Will it feedback?
That's it. Sound reinforcement is conceptually no more difficult than this.
Think about what you do in system design. You estimate the required system level and headroom. Then, you design a loudspeaker system and choose power amplifiers to meet that goal. That answers Question 1.
You complete your loudspeaker system design by making sure you can achieve the required level at every seat in the house. That answers Question 2.
If you calculate ALcons or RASTI or check the signal to noise and direct to reverberant levels for intelligibility, you have answered Question 3.
Finally, by calculating "PAG-NAG" and "NOM" to make sure the system will stay out of feedback, you've answered Question 4.
If you are designing in EASE or Modeler or Ulysses, you may see all of this in graphical form displayed on a 3D model of the room. Nevertheless, you're still answering the same four questions, nothing more.
If you attended the Syn-Aud-Con reunion at NSCA, you had a chance to hobnob with a group of industry veterans who remember when we did all of our design on an HP41 calculator using equations developed by Don Davis and other pioneers. These equations, EPR, PAG-NAG, ALcons and others, gave us answers to Questions 1, 3 and 4.
There were some early graphical tools to answer Question 2. I still have a "Cluster Computer," a transparent sphere developed by John Prohs and distributed by Community Light and Sound. You could sketch a room on this sphere and apply spherical loudspeaker overlays to visualize the coverage of selected loudspeakers. Before that, we had a flat map system called, "Array Perspective," developed by Ted Uzzle at Altec Lansing.
The point of this history lesson is to illustrate that, while today's tools may be more sophisticated, they all still provide answers to the Four Questions, nothing more.
But, what about equalization? What about setting gains and losses? What about reducing hum and noise and distortion? And, what about those FIR and IIR filters and so on? OK, I admit it; there's a fifth question.
Question 5: Does it sound good?
The answer to this question is more complex because "good sound" is application-specific and somewhat subjective. Yet, even Question 5 has a mostly objective answer. Start by designing the system to provide appropriate answers to Questions 1 through 4. What's appropriate? That depends on the application. A speech-only system won't be as loud as a concert reinforcement system (Question 1), but the speech system may have more exacting coverage, intelligibility and feedback requirements (Questions 2, 3 and 4).
Then, plan the system frequency range to be appropriate for the application (wider for music than for voice paging). Choose good quality loudspeakers and use a modest amount of equalization to create a smooth response curve. Strive to minimize hum, noise and distortion. Do all of this well, and you're about 80 percent of the way to a "yes, it sounds good" answer to Question 5. The other 20 percent is the subjective part, the artistry of sound system design and commissioning. Is this subjectivity a bad thing? Well, I, for one, am thankful that I still have something to offer that a machine can't do very well!
And, even considering the subjectivity and the artistry of "good sound," the original Four Questions remain the foundation of sound reinforcement system design. Finally, there are at least two encouraging results from this simple foundation. First, the Four Questions and their answers are almost entirely objective. That means we can set numeric goals for our systems and measure the results to see if we've achieved those goals. This is especially important when you're required to guarantee results for a client.
The second result is that we can use the Four Questions to help our customers understand what we are proposing: "Here's what we intend to achieve, Mr. Customer. And here's how we'll measure the results."
Explain things this way to a customer, and you're way ahead of a competitor who just gives the customer an equipment list and a price. You'll build confidence and trust in your customers' minds and, more than likely, you'll get the order.