Development of a dynamic loudness model including perceptual weights
Final Report Abstract
Loudness is a fundamental auditory sensation, which is correlated with the physical intensity of the sound, but depends on several other stimulus parameters, such as the spectrum and temporal envelope. The sensation loudness is strongly affecting the annoyance of a sound. Hence, it is important to understand all parameters influencing loudness. Current loudness models predict several of these dependencies. One aspect that was not yet included in these loudness models was that different temporal parts of a stimulus are weighted differently when judging the overall loudness of the sound. A prominent effect observed consistently in the temporal loudness weights is the so-called primacy effect, i.e., listeners put a higher weight on the beginning of the sound than to later portions of the sound when judging its loudness. A main aim of this project was to extend our knowledge on this effect by investigating which parameters affect the primacy effect. It provided several important new insights into loudness weights: 1) The primacy effect is hardly affected by the presence of a background noise. 2) The primacy effect is largely independent of the overall level of the sound. 3) Introducing a temporal gap into a sound leads to a second primacy effect for the onset of the sound after this gap. 4) The strength of this second primacy effect depends on the gap duration. 5) This second primacy effect also occurs when a gap is only introduced into one spectral component of a two-component sound. 6) This second primacy effect in the two-component sound is unaffected by the temporal structure of the other component. These results have implications for the modelling of the primacy effect, e.g., result 1) indicates that the primacy effect is also triggered by the onset of a sound when it is not preceded by silence, and results 5) and 6) indicate that temporal weights are applied independently in the different frequency channels. The primacy effect can be predicted by an exponential decay function with the three parameters time constant of the decay, dynamic range, and asymptotic value. In addition, evidence integration processes operating in a decision stage were identified as a potential cause of the primacy effect, providing a link to non-monotonic temporal weights in other sensory judgments and modalities. The project also studied other aspects of the weighting in loudness. For example, several previous studies showed loudness dominance, i.e., that portions of the stimulus with a higher loudness receive a higher weight than softer stimulus portions. For temporal loudness weights, this was replicated in the project and the characteristics of the dominance effect and its potential interaction with the primacy effect were studied. The project also investigated loudness dominance in the spectral domain in normal-hearing and hearing-impaired listeners. In contrast to the temporal domain, in spectral weights, a dominance effect was hardly observed in the listeners of these two groups. Finally, it was investigated if loudness weights are affected by the location of a sound source in space. The hypothesis was that sources which are outside the field of vision may receive higher weights, since for these directions, only the auditory system provides information. In contrast to this hypothesis, the results of the project indicate equal weights for all fundamental directions in space (front, back, right, left, top). One goal of the project was to incorporate a weighting stage in a current loudness model. Simulations with an existing loudness model indicated that it is not possible to just multiply the instantaneous loudness values with the exponential decay function mentioned earlier. The problem is the final stage of the loudness model, which uses the maximum of the loudness-time function as a measure of the overall loudness. A possibility to overcome this problem is to use the mean instead of the maximum. A second challenge for the goal of introducing a weighting function into a loudness model is to have an automatic detection stage that triggers the primacy effect. A first version of such a model was developed but could not be finished before the end of the project.
Publications
- (2018). Evaluation of a model of temporal weights in loudness judgments. Journal of the Acoustical Society of America, 144(2), EL119-EL124
Oberfeld, D., Jung, L., Verhey, J. L., & Hots, J.
(See online at https://doi.org/10.1121/1.5049895) - (2018). Temporal weights in the perception of sound intensity: Effects of sound duration and number of temporal segments. Journal of the Acoustical Society of America, 143(2), 943-953
Oberfeld, D., Hots, J., & Verhey, J. L.
(See online at https://doi.org/10.1121/1.5023686) - (2019). Der Einfluss von Innenohrschwerhörigkeit auf die spektrale Gewichtung bei der Lautheitswahrnehmung. Jahrestagung der Deutschen Gesellschaft für Audiologie, Heidelberg
Hots, J., Oberfeld, D., Fischenich, A., & Verhey, J.L.
(See online at https://dx.doi.org/10.3205/19dga089) - (2019). Parametric measurement of the effects of relative loudness on the relative weights. 23rd International Congress on Acoustics, Aachen
Fischenich, A., Hots, J., Verhey, J. L., & Oberfeld, D.
- (2019). Temporal weights in loudness judgments: A review. 23rd International Congress on Acoustics, Aachen
Oberfeld, D., Fischenich, A., Hots, J., & Verhey, J. L.
- (2019). Temporal weights in loudness: In-vestigation of the effects of background noise and sound level. PloS One, 14(11)
Fischenich, A., Hots, J., Verhey, J., & Oberfeld, D.
(See online at https://doi.org/10.1371/journal.pone.0223075) - (2019). Zeitliche Gewichtung der Lautheitswahrnehmung im Lautheitsmodell. DAGA 2019, Rostock
Hots, J., Oberfeld, D., Fischenich, A., & Verhey, J.L.
- (2020) Räumliche Gewichtung bei der Lautheitswahrnehmung DAGA 2020, Hannover, Germany, ISBN: 978-3-939296-17-1, pp. 808-809, Dega e.V., Berlin
Hots J, Oberfeld D, Fischenich A, Verhey JL
- (2020). The effect of silent gaps on temporal weights in loudness judgments. Hearing Research, 395
Fischenich, A., Hots, J., Verhey, J. L., & Oberfeld, D.
(See online at https://doi.org/10.1016/j.heares.2020.108028) - (2020). What causes the primacy effect in temporal loudness weights? In Proceedings of the e-Forum Acusticum 2021, 3411-3415
Oberfeld, D., Fischenich, A., Ponsot, E., & Verhey, J. L., Hots, J.
(See online at https://doi.org/10.48465/fa.2020.0835) - (2021). Temporal loudness weights are frequency specific. Frontiers in Psychology, 12
Fischenich, A., Hots, J., Verhey, J., & Oberfeld, D.
(See online at https://doi.org/10.3389/fpsyg.2021.588571) - (2021). Temporal loudness weights: Primacy effects, loudness dominance and their interaction. PLOS One, 16, e0261001
Fischenich, A., Hots, J., Verhey, J., Guldan, J., & Oberfeld, D.
(See online at https://doi.org/10.1371/journal.pone.0261001)
