Emerging Technology

Sound System Design and Commissioning

• By Glenn Meeks
• 06/01/14

I have been working through a list of audio- and acoustics-oriented articles. I want to talk about sound systems designed for large-performance or meeting spaces. Those spaces could be an auditorium, cafeteria or even a gymnasium. Good sound system designs start with an empirical, scientific approach, but should always end with listening tests.

Whether your system is designed by an audio professional services firm, electrical engineering firm that uses a local contractor for design or it is a design-build project, the approach is the same.

Ask a trained professional

In today’s world, there are robust applications available that create an acoustical model of the space prior to it actually being built. Input of architectural information into the application is critical and requires training and experience. Therefore, confirm the designer has a number of designs under his or her belt. With the acoustical model available, you may need to work with the architectural team to decrease (and in a few cases for gymnasiums, increase) the amount of reverberation in the space. The modeling program can actually enable you to “hear” music playback and voices in that space, allowing you to make a judgment call on whether you want more or less reverberation. Once the acoustical model is where it needs to be, we move to “speaker mapping.”

Map it out

The concept of “speaker mapping” is to select speakers and place them in the computer model of the space. Speakers are selected based on two criteria: where in the room are we locating the speakers and what does the seating area look like from the speaker location(s)? We want the majority of output from the speakers to land only in the area where people will be sitting. We do not want a large amount of energy from the speaker bouncing off walls or ceiling. Speakers that correctly match up to a broad but shallow space will not work in a narrow and deep space.

Once a speaker is located in the model, a common set of data (all speaker manufacturers use the same standard) is used by the modeling program to create a “map” of how the speaker energy is distributed across the seating area. You want to see that map. Specifically, you want to see multiple maps related to multiple frequencies. Ask for maps starting at 500Hz, and then every octave (1000Hz, 2000Hz) up to 4000Hz. These maps are color-coded, indicating variations in level at that frequency.

Once you understand the map contours, make sure that there is no more than ±1.5dB (maximum difference from loudest to softest area is no more than 3dB); not only on each map but also between maps. Be sure to tell the designer this is what you want before they start the modeling process. Do not be surprised if the requirement causes their eyebrows to rise up. Yes, it is hard to get it that tight, but very possible with today’s speakers. And it really makes a world of difference.

Proper set up is important

Once the system is installed, there are technical processes to make sure things are set up correctly. The first item is simple — make sure every wall or floor connector works and is correctly labeled. The second is called “unity gain,” where the entire electronics chain from input to mixer to input to amplifier is set to 0dB. Unfortunately, most systems I check out have never been properly setup for unity gain. Setting “unity gain” also means basic settings of the DSP (digital signal processor) are correctly set.

Then, using a sound level meter, basic levels between speakers are set using the DSP outputs. Leave the power amplifier inputs wide open; no one can mess up the balanced levels by changing the position of the knob. Now we start equalizing the system. Require use of a dedicated analyzer with high-resolution audio microphone like a “Smart Audio” product or an older TEF (Time Energy Frequency) analyzer (no RTA – real-time analyzers allowed). Once the equalization is set using the equipment, you use Audio CDs (please do not use digital files from a phone or computer) and walk around the seating area while listening.

Understand that the analyzer may indicate things are right, but it doesn’t sound right to someone with a trained ear. Yes, a good audio designer has trained their ears. They can relate what they hear to what is happening technically with the system. Even an experienced professional may take a couple of passes through using equipment and listening to arrive at a good set up for the speakers.

This article originally appeared in the issue of .