Current Issue

May 2012

LOUDSPEAKERS: ARTICULATION, CANCELLATION AND FILTRATION

Guidelines on loudspeaker system design and product design options

By Glenn Ballou

What’s wrong with this approach?
He did not:

1. Say the system will be understandable. (i.e. his proposal did not include Articulation Loss of Consonants (ALCONS) data throughout the listening area)
2. Specify the frequency response.
3. Specify the equivalent acoustic distance (EAD).
4. Specify the coverage pattern and the evenness of the sound pressure level throughout the listening area.
5. Specify the directionality of the sound. (i.e. where does the sound appear to come from?)
6. Specify how many microphones can be on at one time and how close the presenter or instrument must be to the microphone.

There are many things that affect the quality and usefulness of the sound system, but in this article we will only talk about some of the effects caused by the loudspeaker(s). The best loudspeaker in the world improperly installed will probably sound worse than a low-quality loudspeaker properly installed.

If a system is used mainly to amplify the preacher and the choir, a monaural system will almost always work best. As we add loudspeakers in a room, each loudspeaker interacts with the others causing cancellations and comb filtering. For example, sound travels at about 1,130 feet per second, which translates to a 1,000 Hz wave having a wavelength of about 1.13 feet. If two loudspeakers playing the same output material at the same volume are placed the same distance from the listener, the direct sound will be twice as loud as it would if there were only one loudspeaker. (See Figure 1a) However, if the two loudspeakers are placed, or the listener is seated such that one loudspeaker is 1.13 feet farther away from the listener than the second loudspeaker, the direct sound of the 1,000 Hz signal will cancel out, leaving a hole at that frequency. (See Figure 1b).

This same problem exists in most multi-transducer loudspeakers whether they are two-way, three-way, or four-way designs. When the face of each transducer is mounted on a flat surface, two things can occur; 1. Cancellation occurs at the crossover frequencies, 2. The sound each transducer (loudspeaker component) produces will arrive at a different time. That time difference is determined by the distance between each transducer’s acoustic center, or the sound’s point of origin. While most loudspeakers are built with flat fronts, rarely are they built with physically aligned acoustic centers (i.e. the components in the same physical plane). For discussion, let us assume the acoustic center of the low-frequency transducer (woofer) is two inches from the face of the loudspeaker. Mounted to the same flat face is a mid-frequency horn with its acoustic center 14 inches from the face, and a high-frequency transducer with its acoustic center one inch from the face. If the sound from all three transducers originates at the same moment, the high frequencies will arrive at the listener first, followed by the low frequencies then the mid frequencies. Cancellation at the crossover frequencies (overlap points) will produce an effect called comb filtering.

This can be corrected in one of several ways. The first way is to physically align the acoustic centers of each transducer. This can be achieved with special transducer designs that combine the high- and low-, or high- and mid-frequency drivers, such as those built by Tannoy and others, to achieve single point-source alignment and time alignment of multiple frequency bandwidths. Another method requires the front of the loudspeaker cabinet to be a stepped surface. (See Figure 2a) The third, and most common method is to electronically delay the signal of the transducers whose acoustic centers are closest to the listener so they reach the listener at the same time as the signal from the transducer farthest from the listener. (See Figure 2b) This is commonly called “time align” or “signal align”.

This same electronic approach can be used to align loudspeaker clusters. A prime use would be to delay the sound from the loudspeaker mounted one-half way back in the room covering the rear seats so that it arrives at the listener just after the sound from the front loudspeaker. If we do not delay the signal to the rear loudspeaker, the people in the rear of the room would hear it first and then the sound from the front loudspeaker as a delay or an echo. Actually, the sound to the rear loudspeaker should arrive 15 -20 milliseconds (ms) after the sound from the front loudspeaker. This extra delay is called the “Haas Effect” which states that the first sound we hear captures the brain giving the listener the impression that the sound is coming from the front loudspeaker even though the rear loudspeaker is louder and producing the musical clarity and speech intelligibility.

Dispersion characteristics are another important property of the loudspeaker(s). For the clearest and cleanest sound, we want to keep the sound from bouncing wildly off the walls, floors and ceilings. Each acoustic environment requires a specific coverage pattern. For instance, a room that is long and narrow requires a high-Q loudspeaker, or one with a narrow coverage pattern, possibly 40° vertical by 60° horizontal. A room that is very wide, but short, might require a low-Q loudspeaker with 40° vertical by 180° horizontal coverage.

As we add more loudspeakers, our gain before feedback reduces, our articulation will more than likely suffer and our Equivalent Acoustic Distance (EAD) will be reduced. Sound system consultants must first specify certain conditions before he/she can promise a certain sound pressure level (SPL) at the farthest listener. If he states that the SPL will be X dB at the farthest listener within the coverage area of the sound system, he could be held legally accountable for that number. But what were the criteria? Stating a particular EAD would be an ideal criterion. EAD can be explained as the maximum distance a listener with normal hearing can stand from a presenter and clearly and easily hear understand the message acoustically, or without the aid of a sound system. Eight feet is a common design EAD. At any distance beyond the EAD, it will become hard to hear and understand the presenter without a sound system. Therefore, the consultant would design the sound system so that the listener at the farthest seat would hear the presenter through the sound system at the same SPL as the person sitting eight feet away does acoustically (unamplified). If the presenter moves farther from the microphone or talks softer, his SPL at EAD without a sound system and at the farthest seat with a sound system will be diminished and he won’t be heard at either position. But the sound system still meets the design parameters as the consultant cannot control the input to the sound system.

Most presenters and performers require monitors so they can hear themselves. If they cannot hear themselves loudly and clearly, they cannot perform or present at their best. Unfortunately monitors are the biggest cause of feedback. Feedback occurs when the sound from the loudspeaker and the sound from the performer reach the microphone at the same SPL. Rock performers place the microphone about 1/4 inch or less from their mouth and scream into it so that the monitor level coming back to the microphone 4 feet away can be very, very loud! Lately “in-ear monitors” (IEM’s) have been replacing floor monitors, eliminating much of the feedback problems. In-ears also improves the overall sound in the room because the audience hears only the “house” loudspeakers and less sound from the stage. Besides interfering with the house mix, traditional stage monitors feed the microphones with sound that is delayed by the distance between the monitor and the microphone, causing comb filters, etc.

There are many more characteristics of the sound systems and room acoustics which must be taken into consideration when designing a sound system. Good sound system consultants and installers have the proper tools and knowledge to guarantee the results before the system is installed. If they cannot or will not, find one that will! Ask for references, and don’t get fooled by the fast-talking salesman of yesteryear.

Glen Ballou owns Innovative Communications, where he designs and installs sound systems. He is the author/editor of the 1st, 2nd and 3rd editions of the 1552 page “Handbook for Sound Engineers” and also writes for S&VC. Glen can be reached at ici2@mindspring.com.