Related Experiment Videos
A comparison of various nonlinear models of cochlear compression
James M Harte1, Stephen J Elliott, Henry J Rice
1Institute of Sound and Vibration Research, University of Southampton, University Road, Southampton, Hampshire, S017 1BJ, United Kingdom. jmh@isvr.soton.ac.uk
The Journal of the Acoustical Society of America
|July 16, 2005
Summary
Models of the cochlea's basilar membrane response reveal distinct nonlinear behaviors. While some models mimic input-output curves, level-dependent systems exhibit unique multi-valued characteristics not captured by traditional functional modeling.
Area of Science:
- Auditory Neuroscience
- Nonlinear Dynamics
- Bioacoustics
Background:
- The basilar membrane's vibration response to sound exhibits compressive nonlinearity.
- This nonlinear behavior is typically characterized by input-output level curves with varying slopes at different sound levels.
Purpose of the Study:
- To compare different nonlinear system models for simulating the basilar membrane's vibration response.
- To investigate the differences in distortion properties, transient responses, and instantaneous input-output characteristics among these models.
- To assess the suitability of functional modeling techniques for representing these nonlinear systems.
Main Methods:
- Modeling two classes of nonlinear systems: static power-law nonlinearity and level-dependent systems (automatic gain control, Van der Pol oscillator).
- Parameter adjustment to match measured input-output level curves of the basilar membrane.
- Analysis of distortion properties, transient responses, and instantaneous input-output characteristics.
Main Results:
- All considered models can replicate the general input-output level curves of the basilar membrane.
- Static nonlinear models possess single-valued instantaneous characteristics.
- Level-dependent models exhibit multi-valued characteristics with amplitude-dependent slopes.
Conclusions:
- Level-dependent models offer a more nuanced representation of the basilar membrane's nonlinear response.
- The multi-valued, level-dependent nature of these systems challenges the applicability of traditional functional modeling approaches like Wiener or Volterra series.