Discrimination

All distortions are not created equally despite the fact we're told we should strive to have none. For example, harmonic is less objectionable than intermodulation, but if one simply weighs overall distortion numbers without qualifications, or listing all types of distortion, knowledge of the sonic outcome can never be reached simply by these numbers. The problem for potential purchasers relying upon distortion numbers, charts, and graphs to make their decisions is the relevance of each purity contortion and the limited numbers of listed distortion types. And to make matters more complicated, it matters greatly how low levels of distortions are arrived at. If a designer adds enough of the right feedback to achieve astonishingly low distortion levels, how they got there matters. For example, to bring measured distortion down into the -130dB areas often requires an open-loop bandwidth so low as to severely roll of frequencies within the audible range—an act that will in itself form its own type of distortion we do not pay attention to nor measure and publish (FM, SID, TIM, amplitude, phase, group delay, linear, nonlinear, etc.). What might be helpful is for those companies interested in publishing detailed specs of their product to explain the relevance to sound quality for each type of distortion. This might help not only consumers but potentially improve designs (because it would require careful listening tests to create these published details). All distortions are not created equally nor are very many of them paid attention to, let alone published.