Clean amplifier clipping is inaudible until waveforms lose over half their amplitude

Evidence

At the 1987 AES Convention in London, there was a public test of the audibility of amplifier clipping. Although the results were never published, the result was that about 6 dB of clean clipping was rarely audible. I participated in the test and was surprised by the result. Voishvillo (2006, 2007) removed 50% of the waveform amplitude before it became an audible problem. This amount of clipping generates about 20% THD (total harmonic distortion). Interestingly, in comparison he found that this was less annoying than 3% of zero-crossing distortion. — , Toole, Floyd E - Sound reproduction-…, 1792

Categorization

  • CANDY: A delightfully counterintuitive tidbit for any audio enthusiast: heavy clean clipping generates 20% THD and still bothers listeners less than 3% zero-crossing distortion.
  • QUESTION: Why is zero-crossing distortion so far more audible than massive harmonic distortion? The mechanism behind this asymmetry seems unstated here.
  • XREF: connects to: Badly designed loudspeakers cause audible amplifier differences reviewers attribute to the amplifiers — , Toole, Floyd E - Sound reproduction-…, 1769; Audible distortion in quality loudspeakers is rare within intended limits — Toole, Floyd E, Sound reproduction- the acoustics and…, 111; Masking acts within critical bandwidths, leaving distant-frequency sounds unaffected — Eric J. Heller, Why You Hear What You Hear- An…, 2009 (+3 more)

Broader Connections