Publication | Closed Access
A model for the prediction of thresholds, loudness, and partial loudness
657
Citations
0
References
1997
Year
MusicAeroacousticsPsychoacousticsEngineeringNoise PredictionSound QualityLoudness ModelPhoneticsNoiseAcoustical EngineeringStatisticsArchitectural AcousticSpecific LoudnessMiddle EarSpeech PerceptionAuditory ModelingAudiologySignal ProcessingPartial LoudnessSpeech ProcessingAuditory ComputationArtsAuditory System
The model differs from earlier loudness models by its transfer functions, excitation pattern calculations, loudness–excitation relationships, and binaural summation, and assumes a frequency‑independent constant loudness at threshold. The study aims to develop a loudness model for steady sounds that assumes a small but finite loudness at absolute and masked thresholds. The model processes sound through fixed outer‑ and middle‑ear filters, computes an excitation pattern, transforms it to a specific loudness pattern, integrates to obtain overall loudness per ear, and sums loudness across ears. The model accurately reproduces equal‑loudness contours, predicts monaural‑binaural threshold relationships, and explains complex‑sound thresholds and loudness as a function of bandwidth.
A loudness model for steady sounds is described having the following stages: 1) a fixed filter representing transfer through the outer ear; 2) a fixed filter representing transfer through the middle ear; 3) calculation of an excitation pattern from the physical spectrum; 4) transformation of the excitation pattern to a specific loudness pattern; 5) determination of the area under the specific loudness pattern, which gives overall loudness for a given ear; and 6) summation of loudness across ears. The model differs from earlier models in the following areas: 1) the assumed transfer function for the outer and middle ear; 2) the way that excitation patterns are calculated; 3) the way that specific loudness is related to excitation for sounds in quiet and in noise; and (4) the way that binaural loudness is calculated from monaural loudness. The model is based on the assumption that sounds at absolute threshold have a small but finite loudness. This loudness is constant regardless of frequency and spectral content. It is also assumed that a sound at masked threshold has the same loudness as a sound at absolute threshold. The model accounts well for recent measures of equal-loudness contours, which differ from earlier measures because of improved control over bias effects. The model correctly predicts the relation between monaural and binaural threshold and loudness. It also correctly accounts for the threshold and loudness of complex sounds as a function of bandwidth.