From Hydrogenaudio Knowledgebase
Jump to: navigation, search
Claims about audio quality can ultimately only be justified by scientifically reproducible perceptual tests. Agreement on that fact is what makes hydrogenaudio-- and any reasoned discussion about audio-- possible. This principle has been enshrined in the eighth article of hydrogenaudio's Terms of Service:
All members that put forth a statement concerning subjective sound quality, must -- to the best of their ability -- provide objective support for their claims. Acceptable means of support are double blind listening tests (ABX or ABC/HR) demonstrating that the member can discern a difference perceptually, together with a test sample to allow others to reproduce their findings. Graphs, non-blind listening tests, waveform difference comparisons, and so on, are not acceptable means of providing support.

A test of audio quality is scientific only if it is blind. This eliminates the effects of pre-existing opinions, the placebo effect, and other psychological biases so we can consider the audible differences.

Test results are convincing only when they have been shown to be reproducible. A single blind test, while a useful data point, doesn't constitute convincing evidence. Even a single listener's responses to perceptual tests will vary from moment to moment, so a single result could be "due to chance" rather than showing a real quality difference. Statistical methods-- both frequentist and Bayesian-- can help us recognize convincing evidence when we combine the results of multiple observations.

Without agreement to subject all quality claims to scientifically reproducible testing, conversation about audio rapidly devolves into shouting matches about pre-existing opinions and biases, detailed "analyses" of differences in the waveform that are entirely irrelevant to human perception, and attempts to browbeat others into submission by the sheer repetition of impressive-sounding mumbo-jumbo "terms of art" which are completely subjective and therefore communicate nothing.