Originally Posted by ouchia
let's also not forget that measurement devices currently used are not as sensitive as the human ear. assuming that you're listening to your gear with a human ear that is...
That's only partially true. It depends on what we are trying to "measure". For amplitude, frequency, stereo separation, THD, and all the common measurements given, measurement devices are millions of times more sensitive than the ear.
The smallest change in amplitude that the ear can detect is about 0.3dB for a pure tone. However for most people this is closer to 0.5 to 1.0dB, or a 10% change.
The smallest detectable change in frequency we can detect is 0.2% in the 500Hz-2KHz band. In what can be expressed in percentages, this is the most sensitive the ear gets.
The smallest amount of harmonic distortion the ear can detect is generally going to be about 1% for low-order harmonics, or 0.3% for crossover effects. A telephone is 10% THD, and most solid state amps are 0.005% or better in THD. But THD is of limited use in predicting how an amp will sound subjectively. Intermodulation distortion is a more important figure (and less often given), however THD graphs can give the observer a better idea of where and why a non-linearity exists.
Where the ear excels...the human ear is at it's best in placing sounds in a 3D space. We can determine the location of a sound much better than a machine. The human ear is also superior to machines in recognizing patterns such as speech. This makes sense as these abilities provide an evolutionary advantage. We could probably do an excellent job analyzing binaural recordings in comparing them to the setting in which they were recorded.