|[ Team LiB ]|
As a human sensation, sound is an analog phenomenon. It has two primary characteristics—loudness (or amplitude) and frequency—that vary over a wide range with an infinite range of variations between its limits. Sounds can be loud or soft or any gradation in between. Frequencies can be low or high or anything in between.
Sound frequencies are measured in hertz, just like the frequencies of computer clocks or radio signals. The range of frequencies that a human being can hear depends on age and sex. Younger, female ears generally have wider ranges than older, male ears—maybe that's why some older adults miss the nuances of rap music. Maybe not. In any event, most sources ignore individual differences and list the range of human hearing as being from 20 hertz to 15,000 hertz (or as high as 20,000 hertz if your ears are particularly good).
Lower frequencies correspond to bass notes in music and the thumps of explosions in theatrical special effects such as exploding spaceships and particularly tenacious indigestion. High frequencies correspond to the "treble" in music, the bright, even harsh sounds that comprise overtones in music—the brightness of strings, the tinkle of bells, the sharp edge of crashing glass—as well as hissy sounds, such as sibilants in speech, the rush of waterfalls, and overall background noise.
Low frequencies have long wavelengths, in the range of ten feet (three meters) for middle bass notes. The long wavelengths allow low frequencies to easily bend around objects and, from a single speaker, permeate a room. Moreover, human hearing is not directionally sensitive at low frequencies. You cannot easily localize a low frequency source. Acoustical designers exploit this characteristic of low frequencies when they design low-frequency loudspeakers. For example, because you cannot distinguish the locations of individual low-frequency sources, a single speaker called a subwoofer is sufficient for very low frequencies, even in stereo and multichannel sound systems.
High frequencies have short wavelengths, measured in inches or fractions of an inch (or centimeters). They can easily be blocked or reflected by even small objects. Human hearing is acutely sensitive to the location of higher frequency sources.
Amplitude describes the strength or power of the sound. The amplitude of sound traveling through the air is usually expressed as its sound pressure level (SPL). The threshold of human hearing is about 0.0002 microbars—which means a pressure change of 1/5,000,000,000th (one five-billionth) of normal atmospheric pressure. In other word, the ear is a sensitive detector of pressure changes. Were it any more sensitive, you might hear the clink of Brownian motion as dust particles ricochet through the air.
In audio systems, electrical signals take the place of sound pressure waves. These signals retain the characteristic frequencies of the original sounds but their amplitude refers to variations in electrical strength. Usually the voltage in an audio system represents the amplitude of pressure of the original sound waves.
A term that you'll usually see engineers use in measuring amplitude loudness is the decibel (dB). Although the primary unit is actually the bel, named after Alexander Graham Bell, the inventor of the hydrofoil (and yes, the telephone), engineers find units of one-tenth that quantity to be more manageable.
The decibel in itself represents not a true measuring unit but a relationship between two measurements. The bel is the ratio between two powers expressed as a logarithm. For example, a loud sound source may have an acoustic power of one watt, whereas a somewhat softer source may only generate one milliwatt of power, a ratio of 1000:1. The logarithm of 1000 is 3, so the relationship is 3 bels or 30 decibels—one watt is 30 decibels louder than one milliwatt.
In addition to reducing power relationships to manageable numbers, decibels also approximately coincide with the way we hear sounds. Human hearing is also logarithmic, which means that something twice as loud to the human ear does not involve twice the power. For most people, for one sound to appear twice as loud as another, it must have ten times the power. Expressed as dB, this change is an increase in a level of three dB because the logarithm of 10 is 0.3, so the relationship is 0.3 bels or 3 decibels.
Engineers also use the decibel to compare sound pressures and voltages. Because sound pressures represent power, the math is the same as for electrical power. But for voltages, the decibel relationship is different. Power varies as the square of the voltage, so a doubling in voltage results in four times more power (four is two squared). With logarithms, increasing a quantity by the power of two doubles a logarithm. Consequently, a doubling of voltage represents not a 3 dB change but one of 6 dB.
Most commonly you'll see dB used to describe signal-to-noise ratios and frequency response tolerances. The unit is apt in these circumstances because it is used to reflect relationships between units. Sometimes, however, people express loudness or signal levels as a given number of "dB." This usage is incorrect and meaningless because it lacks a reference value. When the reference unit is understood or specified, however, dB measurements are useful.
Any unit may be used as a reference for measurements expressed in dB. Several of these have common abbreviations, as listed in Table 25.1.
In sound systems, the dBm system is most common. The electronic industry adopted (in May, 1939) a power level of one milliwatt in a 600 ohm circuit as the standard reference for zero dBm.
You'll also encounter references to volume units (VU), as measured by the classic VU meter. A true VU meter is strictly defined, and the zero it indicates reflects a level 4 dB above 0 dBm. The meters on tape and cassette recorders have VU designations but are not, strictly speaking, VU meters. They are usually referenced to a specific recording level on the tape. Similarly, meters on radio transmitters are often calibrated in VU, but zero corresponds not to the input line level but the output modulation level—0 VU usually means 100-percent modulation.
No one is perfect, and neither are audio amplifiers. All analog audio amplifiers add subtle defects to the sound called distortion. In effect, distortion adds unwanted signals—the defects—to the desired audio. The most common way to express distortion is the ratio between the unwanted and wanted signals expressed as a percentage. In modern low-level circuits, the level of added distortion is vanishingly small, hundredths of a percent or less, and is often lost in the noise polluting the signal. Only if your hearing is particularly acute might you be able to detect the additions.
In power amplifiers, the circuits that produce the high-level signals used to operate nonpowered loudspeakers, the addition of distortion can rise to levels that are not only noticeable but also objectionable, from one-tenth to several percent. Power amplifier distortion also tends to increase as the level increases—as you turn the volume up.
Better amplifiers produce less distortion. This basic fact has a direct repercussion when it comes to getting the best sound quality from a computer. The soundboards typically found in computers produce a lot of distortion. You can often get appreciably better audio quality by plugging your stereo system (which is designed for low distortion even at high power levels) into the low-level outputs of your soundboard so that the audio signal avoids the soundboard's own power amplifier. In fact, better-quality soundboards often lack power amplifiers in the recognition that any included circuitry is apt to be of lower sound quality due to the restrictions of installing them on a circuit board with limited power supply capacity.
|[ Team LiB ]|