Audio Basics: A Is For Ampere Page 10
Ears process sounds in three ways: by loudness, frequency, and time difference. Loudness is a function of the range (above and below atmospheric) through which the air pressure is fluctuating, but our ears don't respond to intensity in a linear fashion. Rather, we hear volume changes incrementally. A doubling of sound pressure is perceived as being the same amount of volume increase, whether the sound was very soft or very loud to begin with. For this reason, loudness (called sound-pressure level or spl) is measured on a logarithmic deciBel scale, which better reflects how we hear volume changes. The reference point for spl---0dB---is the softest sound we can hear; 120dB verges on physical discomfort. The total range represents a ratio of 1 to 1,000,000.
Low frequencies, from 20 to around 250Hz, are actually traced by variations in the density of neural impulses, ranging from few impulses to many thousands per second. Above 250Hz, frequencies are perceived according to the positions in the cochlea occupied by the nerve-endings they stimulate.
The time element in hearing has to do with the delay between sounds reaching one ear and then the other. It is still not certain how timing information is processed, but it appears related to the system's ability to delay left- or right-ear input until synchronism is achieved, and then to "calculate" from the amount of correction the direction from which a sound is coming, relative to the angle of the head to the source. Any sources which are not "synchronized" are heard as having random or unspecific directionality, a large variety of which together are perceived as having the quality of space or ambience. (Note that you cannot "concentrate" on more than one stereo image at a time. All others have very vague specificity.)
Before it was even called high-end audio, the pursuit of realistic sound reproduction started with the assumption that the more nearly perfect the reproducing equipment, the more nearly perfect (that is, the more hi-fi) the reproduction would be. For many years, objectivity tracked subjectivity; the better something measured, the better it sounded. But during the late 1950s, it started to become evident that critical listeners were hearing differences that weren't being measured. Stereophile was founded in 1962 because, until then, no publication acknowledged the self-evident fact that music-lovers bought components to listen to, not to measure, so the ear, not the meter, was the ultimate arbiter of audio quality.
The audio engineering community still refuses to admit this, although it has nonetheless accepted some new testing techniques which correlate better with subjective observations than the simplistic frequency-response-plus-THD measurements of the '60s. (Impedance curves and harmonic-by-harmonic analyses, to mention two.) As a result, components today are much better (by any criteria) than they used to be, but as long as transducers are mechanical, they will continue to be subject to a host of resonance-induced colorations which, by their very nature, are so complex that it is still impossible to predict, by any objective means, how a given loudspeaker system will sound in a given room with a given amplifier playing a given recording. The fact that this variability is clearly audible to those who will listen, and is supremely amenable to user modification (called "tweaking"), is the whole thrust behind high-end audio, and the reason it is such a rewarding hobby for those who appreciate the differences they hear.
There, in a lengthy nutshell, are about all the boring basics you'll ever need to know in order to understand how audio works. We'll look at some of them in more detail in future articles about specific components, but if you've absorbed most of this stuff, you'll already have a much better comprehension of what your system is doing (and why) than you ever had before. Happy tweaking.