Low-Efficiency Speaker Systems.

We should first talk about the term "sound pressure level" (abbreviated "SPL") which your ears interpret as "loudness" or "volume." SPL is almost always expressed in "decibels," abbreviated "dB." dB's are thrown around a lot, without much understanding. When you talk dB's you're always talking the difference between two quantities. For example, "100 dB SPL" means a sound pressure level 100 dB above a 0 dB point set at the lowest sound pressure level discernible by the average human ear. You're also talking differences when you say one sound is 3 dB louder than another sound. A 3 dB difference can be heard by most listeners but it certainly won't knock your socks off. It takes something of the order of a 10 dB difference in average SPL to be perceived as a doubling (or halving) of loudness. Yet doubling amplifier power, or adding a second speaker system, gives only a 3 dB increase in output. The same result can be obtained by using a speaker with twice as much efficiency. Now we can talk about "efficiency." Speaker system efficiency is the amount of sound a speaker system is able to put out for a certain amount of electrical (audio) signal fed in. From this, you can see that high efficiency is good. It means that for a given amount of amplifier power you can get more sound from the speaker system. Efficiency is properly presented by a percentage. For example, a high-efficiency, direct-radiator speaker system (such as the E-V S15-3) approaches 5% efficiency. A good compression driver/horn combination (such as the E-V DH 1506/HR60) approaches 25% efficiency. What all this means is that high-efficiency speaker systems can produce high sound pressure levels. Well designed speaker systems usually incorporate high efficiency as one of their design goals. All Electro-~ice systems from the S12-2 on up are designed to give the highest efficiency possible for their size and type. High efficiency means you can obtain the sound pressure level you want in a room without distorting the power amplifier, which is the next subject.

Not Enough Amplifier Power.

Before any speaker system can perform to its highest potential, it must be connected to an adequate amplifier, especially one with sufficient headroom. We don't mean how small a foreign car's interior is! Headroom is the amount of reserve level capability that the amplifier has above the long-term average level your ears hear as "loudness." In live music, 10 dB peaks above the average - of a few milliseconds duration - are common and are continuously going through the system. If the peaks can't get through, the sound will still be as loud but it will sound rough and distorted. This means that if you are playing at a 10-watt average level, you will need a 100-watt power amplifier to pass the peak levels (10 dB higher) without clipping (distorting) the amplifier. When the amplifier does go into clipping, you will know immediately because you will hear it through the speakers. Many times people say their speakers sound bad at medium-to-high levels when in reality it is their amplifier. The speaker has no choice but to reproduce the signal being fed to it whether it is clean or distorted.

This amplifier clipping is also a common cause of speaker failure. When clipping occurs, high-level high frequencies are produced which usually overpower tweeters and midrange speakers and result in smoke and no sound! Therefore, you need to be certain that the amplifier you use has enough power to give plenty of reserve over the average needed to give the desired sound pressure level.

Here's an example to help make the efficiency-versus-power issue more clear. Let's say you are running a 250-watt amp into speakers that are 2_% efficient. You're operating at 25-watts average, so there is headroom for the peaks. But it isn't loud enough (SPL too low). You could go to a 500-watt amp and get 3 dB more,' but that's expensive and still not very impressive in additional loudness. You could also go to a 5%-efficient speaker to get the same 3 dB; or you could use a very efficient speaker (maybe 25%) and get really loud but still be clean and have headroom. The point is, speaker efficiency is at least as important as amplifier power

Poor Frequency Response.

Assuming you are using a high-quality microphone, mixer, and amplifier, your speaker system may have poor frequency response. Frequency response is the way a speaker responds (in dB) to a constant input signal swept over the audible frequency range from low bass to the highest treble. Speakers that have varying frequency response will also vary when music is being played through them. This causes the speaker to produce an unnatural, "colored" sound. A speaker that exhibits a flat frequency response over the frequency range it is intended to be used for will sound more natural than one which varies up and down over the same frequency range. A flat response is also desirable to reduce feedback. If a speaker has a large peak in response, the microphone may respond to that peak first and feedback will occur at the frequency where the peak is.

Figure 1 shows how sound pressure level varies with frequency at a specified distance in front of the speaker under anechoic (non-reflecting) conditions, which is similar to being outdoors where there are no walls or ceiling to reflect and modify the sound. The best all-around result will come from the speaker with "flat" response. If you want the sound shaped in some way, do it with an equalizer, or the EQ on your board, where you control it.

Highs Miss Half Your Audience.

Let's say your system uses only a 12-inch cone speaker mounted in a box. Take a look at Figure 2.

At low frequencies sound is dispersed over a very wide angle (in fact, almost omni directional). This is because the cone is small in comparison to the wavelength of low frequencies. Wavelength is velocity of sound in air (1130 feet per second) divided by the frequency. So, for example, the wavelength at 50Hz is 22.6 feet. That is quite a bit larger than a 12-inch speaker. As you go up in frequency, the 12-inch speaker becomes larger than the wavelength and a phenomenon called beaming starts to occur. This is where, instead of having very wide dispersion or coverage angle, you have increasingly narrow coverage This is why the listeners at the side of the room sometimes can't hear the high frequencies. The sound is dull and unintelligible .Therefore, coverage angle (dispersion) over the range of frequencies involved is an important consideration in speaker system design.

A speaker's dispersion of sound as the listener moves to various angles off the speaker axis is typically shown in a polar response graph such as in Figure 3. Measurements are usually made in both the horizontal (side-to-side) and vertical (up-and-down) planes. Both are shown in Figure 3.

A typical approach is to feed the speaker a test signal containing all of the frequencies in an octave, like 2400 to 4800 Hz. This avoids the confusing variations of single-frequency measurements. Since the test signal contains all the frequencies in the octave of interest, it has no definable pitch or musicality but, instead, sounds something like the between-stations noise on an FM receiver. Therefore, the signal is called "random noise." The loudness (SPL) of this noise is measured at all points around the speaker, at a constant distance away, and the level is recorded on the polar graph. Note in the example shown in Figure 3 that, in both the horizontal and vertical planes, the 2400-4800 Hz frequencies are about 10 dB louder in front of the speaker than 60 degrees off to the side. Remember, a 10 dB difference in SPL is perceived as "twice as loud." So, with some frequencies "half as loud," some "nearly as loud," and some nearly gone, it is little wonder that people at the side of the room hear poor. muddy sound from a speaker that has not really considered uniform dispersion in the design.

It is convenient to say a speaker has a certain coverage angle (90° etc.), but if you don't know the dispersion or coverage angle of the speaker for each octave band, you can be misled. Some manufacturers say their speaker has 90° dispersion and that's that. This would be sufficient if it were true for all frequencies. However, in the real world even the best loud speakers only approach this goal. Most differ greatly over their frequency range. Therefore, Electro-Voice supplies not only polar responses but also beamwidth-versus-frequency graphs for most speaker products, as shown in Figure 4. From such a graph you can determine coverage angle for frequencies important to you. In Figure 4 and in our engineering data sheets, we've defined the coverage angle in each octave band as the angle included by the points on the polar response where speaker output is 6 dB below the on-axis response. Although no absolute standard exists, this definition of "coverage angle" or "beamwidth" is often used.

Uniform dispersion is one of the most important and most neglected characteristics of a speaker system. Electro-Voice provides coverage-angle data in the form of beamwidth and polar response on all its products, so you can design a system that will put the sound where you want it. To design a speaker system that possesses uniform dispersion, special components are used. For example, the E-V 515-3 is a three-way, full-range system. It has a 15-inch low-frequency driver which is only used up to 600 Hertz, so its coverage won't get too beamy or narrow. Then, from 600 Hertz to 4000 Hertz, a small 6 1/2-inch cone midrange is used, and from 4000 to 18,000 Hertz a wide-angle, constant-dispersion horn tweeter is utilized. (Some horns have a beaming problem similar to a 12-inch cone speaker, for example.) Using separate components designed for operation over their portion of the audio spectrum instead of using just one speaker will generally yield superior overall performance where the application calls for reproducing the full frequency range. Knowing something about the dispersion angle can help you select speakers for your application. Speakers should be directed to cover the listeners. Viewing the listening area from a desired speaker location, determine what dispersion angle would be needed to adequately cover the listeners without spilling over to the walls in both the horizontal and vertical planes. Once these angles are determined, the correct speaker can be found by consulting E-V engineering data sheets and catalogs.