rhc@hou2a.UUCP (06/17/83)
Care must be taken when discussing audio component performance solely
in terms of frequency response and Fourier analysis. Fourier analysis,
which gives us the concept of frequency response, is a steady state
analysis of linear systems. Music is transient by nature; and audio
components, by virtue of the fact that they have harmonic and
intermodulation distortion, are not linear. Even for a linear system,
the frequency response must be known over the infinite frequency range
for the system to be uniquely described and the transient response
to be known.
The implication here is that specifying the frequency response
of an audio system within a finite frequency range is far from a complete
description of the system. Stating the frequency response over the
range of, say 20 to 20,000 hertz, does not give a clear indication of
how the system will respond to dynamic music, even music with most
of its steady state components within that frequency range. In practice,
this is why different audio components with similar frequency response
curves will have strikingly different levels of claritly, presence, etc.,
on identical musical passages.