Sounds with steep spectral peaks evoke stronger pitch than sounds with narrow spectral notches.

Evidence

When comparing the spectra of peaked ripple noise displayed in Fig. 5.17 and of comb-filtered noise displayed in Fig. 5.24 (sound 9), the two sounds show a figure-ground relation: peaked ripple noise shows sharp spectral peaks and broad spectral valleys, whereas comb-filtered noise shows broad spectral peaks and narrow spectral valleys. Because the pitch strength of peaked ripple noise is larger by a factor of more than 5 than that of comb-filtered noise, it can be assumed that large pitch strength is produced by sounds containing narrow bands with steep spectral slopes rather than by narrow gaps. — Professor Dr.-Ing. Hugo Fastl, Professor Dr.-Ing. Eberhard, Psychoacoustics- Facts and Models, 830

Categorization

Category: Contrarian or Surprising Claims

  • DEEPTHINKING: A figure-ground asymmetry in pitch perception: steep ‘positive’ spectral features carry pitch far better than the complementary steep ‘negative’ gaps — a non-obvious asymmetry that bears on how the auditory system codes spectral edges.
  • XREF: Echoes of figure-ground processing in visual and multisensory perception; also connects to Gabor-filter and spectral-edge coding models in auditory neuroscience.