Limits...
The dynamic range paradox: a central auditory model of intensity change detection.

Simpson AJ, Reiss JD - PLoS ONE (2013)

Bottom Line: However, while loudness grows as intensity is increased, improvement in intensity discrimination performance does not follow the same trend and so dynamic range estimations of the underlying neural signal from loudness data contradict estimations based on intensity just-noticeable difference (JND) data.From the modeling, the following central adaptation parameters are derived; central dynamic range of 0.215 sones, 95% central normalization, and a central loudness JND constant of 5.5×10(-5) sones per ms.Through our findings, we argue that loudness reflects peripheral neural coding, and the intensity JND reflects central neural coding.

View Article: PubMed Central - PubMed

Affiliation: Centre for Digital Music, Queen Mary University of London, London, United Kingdom. andy.simpson@eecs.qmul.ac.uk

ABSTRACT
In this paper we use empirical loudness modeling to explore a perceptual sub-category of the dynamic range problem of auditory neuroscience. Humans are able to reliably report perceived intensity (loudness), and discriminate fine intensity differences, over a very large dynamic range. It is usually assumed that loudness and intensity change detection operate upon the same neural signal, and that intensity change detection may be predicted from loudness data and vice versa. However, while loudness grows as intensity is increased, improvement in intensity discrimination performance does not follow the same trend and so dynamic range estimations of the underlying neural signal from loudness data contradict estimations based on intensity just-noticeable difference (JND) data. In order to account for this apparent paradox we draw on recent advances in auditory neuroscience. We test the hypothesis that a central model, featuring central adaptation to the mean loudness level and operating on the detection of maximum central-loudness rate of change, can account for the paradoxical data. We use numerical optimization to find adaptation parameters that fit data for continuous-pedestal intensity change detection over a wide dynamic range. The optimized model is tested on a selection of equivalent pseudo-continuous intensity change detection data. We also report a supplementary experiment which confirms the modeling assumption that the detection process may be modeled as rate-of-change. Data are obtained from a listening test (N = 10) using linearly ramped increment-decrement envelopes applied to pseudo-continuous noise with an overall level of 33 dB SPL. Increments with half-ramp durations between 5 and 50,000 ms are used. The intensity JND is shown to increase towards long duration ramps (p<10(-6)). From the modeling, the following central adaptation parameters are derived; central dynamic range of 0.215 sones, 95% central normalization, and a central loudness JND constant of 5.5×10(-5) sones per ms. Through our findings, we argue that loudness reflects peripheral neural coding, and the intensity JND reflects central neural coding.

Show MeSH
Simulation of pseudo continuous data; miscellaneous.Predictions of the central excitation pattern model (dashed grey line) for various data; A group mean thresholds of the supplementary experiment (circles); noise pedestals with up-down ramps, at half-ramp durations of 5, 10, 100, 1000, 10000 and 50000 ms and at an overall listening level of 33 dB SPL (rms). Error bars represent 95% confidence intervals. The trends shown in the data are significant (p = 9.55×10−8, Friedman Rank Sum Test). B Just-noticeable difference for envelope modulation of a 1 kHz tone, as a function of beat frequency, produced with the method of beats by Riesz for a listening level of 50 dB SL. C Just-noticeable difference for detection of symmetrical, linearly-ramped increments in 20-dB spectrum-level noise pedestals, as a function of half-ramp duration (one-sided) - averaged data of Plack et al. (circles). D Just-noticeable difference for increment detection in 477 Hz pedestals, as a function of increment duration at a peak level of 60 dB SPL - averaged data of Gallun and Hafter (circles). E, F JND for increment and decrement detection in 4 kHz pedestals respectively, as a function of duration at a listening level of 55 dB SPL - averaged data of Oxenham for 500 ms pedestals presented in quiet (circles), 0 dB (triangles) and 20 dB (squares) spectrum level noise.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3585315&req=5

pone-0057497-g005: Simulation of pseudo continuous data; miscellaneous.Predictions of the central excitation pattern model (dashed grey line) for various data; A group mean thresholds of the supplementary experiment (circles); noise pedestals with up-down ramps, at half-ramp durations of 5, 10, 100, 1000, 10000 and 50000 ms and at an overall listening level of 33 dB SPL (rms). Error bars represent 95% confidence intervals. The trends shown in the data are significant (p = 9.55×10−8, Friedman Rank Sum Test). B Just-noticeable difference for envelope modulation of a 1 kHz tone, as a function of beat frequency, produced with the method of beats by Riesz for a listening level of 50 dB SL. C Just-noticeable difference for detection of symmetrical, linearly-ramped increments in 20-dB spectrum-level noise pedestals, as a function of half-ramp duration (one-sided) - averaged data of Plack et al. (circles). D Just-noticeable difference for increment detection in 477 Hz pedestals, as a function of increment duration at a peak level of 60 dB SPL - averaged data of Gallun and Hafter (circles). E, F JND for increment and decrement detection in 4 kHz pedestals respectively, as a function of duration at a listening level of 55 dB SPL - averaged data of Oxenham for 500 ms pedestals presented in quiet (circles), 0 dB (triangles) and 20 dB (squares) spectrum level noise.

Mentions: Fig. 5a shows the results of the supplementary experiment (see Experiment S1). Group mean thresholds for the 10 listeners are given, including error bars representing 95% confidence intervals. The trends shown in the data are significant (p = 9.55×10−8, Friedman Rank Sum Test). The results, plotted on a logarithmic (time) scale, show symmetry about the half-ramp of 100-ms ‘best detection point’ which appears equivalent to that shown around 3–4 Hz by Riesz (Fig. 5b). Furthermore, the results confirm Riesz’s general finding that slow ramps are hard to detect. It should be noted that short-term (trace) memory [30] may play a role in the results at very long ramps (i.e., >4 seconds), in that the listener is forced to assess the intensity change within the short-term memory window (trace).


The dynamic range paradox: a central auditory model of intensity change detection.

Simpson AJ, Reiss JD - PLoS ONE (2013)

Simulation of pseudo continuous data; miscellaneous.Predictions of the central excitation pattern model (dashed grey line) for various data; A group mean thresholds of the supplementary experiment (circles); noise pedestals with up-down ramps, at half-ramp durations of 5, 10, 100, 1000, 10000 and 50000 ms and at an overall listening level of 33 dB SPL (rms). Error bars represent 95% confidence intervals. The trends shown in the data are significant (p = 9.55×10−8, Friedman Rank Sum Test). B Just-noticeable difference for envelope modulation of a 1 kHz tone, as a function of beat frequency, produced with the method of beats by Riesz for a listening level of 50 dB SL. C Just-noticeable difference for detection of symmetrical, linearly-ramped increments in 20-dB spectrum-level noise pedestals, as a function of half-ramp duration (one-sided) - averaged data of Plack et al. (circles). D Just-noticeable difference for increment detection in 477 Hz pedestals, as a function of increment duration at a peak level of 60 dB SPL - averaged data of Gallun and Hafter (circles). E, F JND for increment and decrement detection in 4 kHz pedestals respectively, as a function of duration at a listening level of 55 dB SPL - averaged data of Oxenham for 500 ms pedestals presented in quiet (circles), 0 dB (triangles) and 20 dB (squares) spectrum level noise.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3585315&req=5

pone-0057497-g005: Simulation of pseudo continuous data; miscellaneous.Predictions of the central excitation pattern model (dashed grey line) for various data; A group mean thresholds of the supplementary experiment (circles); noise pedestals with up-down ramps, at half-ramp durations of 5, 10, 100, 1000, 10000 and 50000 ms and at an overall listening level of 33 dB SPL (rms). Error bars represent 95% confidence intervals. The trends shown in the data are significant (p = 9.55×10−8, Friedman Rank Sum Test). B Just-noticeable difference for envelope modulation of a 1 kHz tone, as a function of beat frequency, produced with the method of beats by Riesz for a listening level of 50 dB SL. C Just-noticeable difference for detection of symmetrical, linearly-ramped increments in 20-dB spectrum-level noise pedestals, as a function of half-ramp duration (one-sided) - averaged data of Plack et al. (circles). D Just-noticeable difference for increment detection in 477 Hz pedestals, as a function of increment duration at a peak level of 60 dB SPL - averaged data of Gallun and Hafter (circles). E, F JND for increment and decrement detection in 4 kHz pedestals respectively, as a function of duration at a listening level of 55 dB SPL - averaged data of Oxenham for 500 ms pedestals presented in quiet (circles), 0 dB (triangles) and 20 dB (squares) spectrum level noise.
Mentions: Fig. 5a shows the results of the supplementary experiment (see Experiment S1). Group mean thresholds for the 10 listeners are given, including error bars representing 95% confidence intervals. The trends shown in the data are significant (p = 9.55×10−8, Friedman Rank Sum Test). The results, plotted on a logarithmic (time) scale, show symmetry about the half-ramp of 100-ms ‘best detection point’ which appears equivalent to that shown around 3–4 Hz by Riesz (Fig. 5b). Furthermore, the results confirm Riesz’s general finding that slow ramps are hard to detect. It should be noted that short-term (trace) memory [30] may play a role in the results at very long ramps (i.e., >4 seconds), in that the listener is forced to assess the intensity change within the short-term memory window (trace).

Bottom Line: However, while loudness grows as intensity is increased, improvement in intensity discrimination performance does not follow the same trend and so dynamic range estimations of the underlying neural signal from loudness data contradict estimations based on intensity just-noticeable difference (JND) data.From the modeling, the following central adaptation parameters are derived; central dynamic range of 0.215 sones, 95% central normalization, and a central loudness JND constant of 5.5×10(-5) sones per ms.Through our findings, we argue that loudness reflects peripheral neural coding, and the intensity JND reflects central neural coding.

View Article: PubMed Central - PubMed

Affiliation: Centre for Digital Music, Queen Mary University of London, London, United Kingdom. andy.simpson@eecs.qmul.ac.uk

ABSTRACT
In this paper we use empirical loudness modeling to explore a perceptual sub-category of the dynamic range problem of auditory neuroscience. Humans are able to reliably report perceived intensity (loudness), and discriminate fine intensity differences, over a very large dynamic range. It is usually assumed that loudness and intensity change detection operate upon the same neural signal, and that intensity change detection may be predicted from loudness data and vice versa. However, while loudness grows as intensity is increased, improvement in intensity discrimination performance does not follow the same trend and so dynamic range estimations of the underlying neural signal from loudness data contradict estimations based on intensity just-noticeable difference (JND) data. In order to account for this apparent paradox we draw on recent advances in auditory neuroscience. We test the hypothesis that a central model, featuring central adaptation to the mean loudness level and operating on the detection of maximum central-loudness rate of change, can account for the paradoxical data. We use numerical optimization to find adaptation parameters that fit data for continuous-pedestal intensity change detection over a wide dynamic range. The optimized model is tested on a selection of equivalent pseudo-continuous intensity change detection data. We also report a supplementary experiment which confirms the modeling assumption that the detection process may be modeled as rate-of-change. Data are obtained from a listening test (N = 10) using linearly ramped increment-decrement envelopes applied to pseudo-continuous noise with an overall level of 33 dB SPL. Increments with half-ramp durations between 5 and 50,000 ms are used. The intensity JND is shown to increase towards long duration ramps (p<10(-6)). From the modeling, the following central adaptation parameters are derived; central dynamic range of 0.215 sones, 95% central normalization, and a central loudness JND constant of 5.5×10(-5) sones per ms. Through our findings, we argue that loudness reflects peripheral neural coding, and the intensity JND reflects central neural coding.

Show MeSH