AES Section Meeting Reports

Melbourne - September 12, 2011

Meeting Topic:

Moderator Name:

Speaker Name:

Other business or activities at the meeting:

Meeting Location:


Dr Staples introduced the work of Jim Fulton, hearing researcher, describing how the inner ear converts sound into neural impulses.
After describing the basic arrangement of the components of the ear, the focus changed to concentrate on the physical and neurological construction of the inner ear — the Cochlea and associated neural connections.
The presentation showed that in cross-section, the rigidly attached supports for the Basilar Membrane to the sides of the Cochlea, and the membrane rigidly braced longitudinally by the Organ of Corti constrain the Basilar Membrane from vibrating as traditionally described. Instead, it is likely to provide a structurally rigid base on which the hearing transducers are mounted.
Fulton describes a gel-like substance on the underside of the Tectorial Membrane, another rigid structure parallel to the Basilar Membrane, which carries a surface acoustic wave and acts like an acoustical waveguide. This gel SAW waveguide called "Henson's Stripe" is in contact with the Inner Hair Cells. Henson's stripe is also adjacent to a flatter plane surface of the gel in contact with the Outer Hair Cells.
The curvature of the cochlea provides the frequency selectivity observed in the Cochlea. Marcatili modelled the dispersive waveguide in optical fibres: Fulton argues that a similar mechanism to the dispersive Marcatili filter is how Henson's stripe performs the frequency selectivity. As the frequency reaches a "critical" frequency, the wave at that narrowly defined frequency leaves the waveguide to travel across the nearby membrane and intersects the chevrons of Outer Hair Cells.
The Inner Hair Cells in contact with Henson's stripe sense a low-pass filtered temporal loudness signal component, while the very narrow band signals from the dispersive filter feed the Outer Hair Cells.
The hair cells transducer the mechanical vibrations in the gel into an electrical signal, which is then amplified by biological "transistors" that Fulton calls "Activas" formed at the junction between cells. Fulton's "Electrolytic Theory" of the neuron is different from the traditional biochemical model, and may be controversial in the traditional context.
Dr. Staples has in part modelled the circuit topologies suggested by Fulton between the cochlea and the ganglion cells of the audio nerve. This model demonstrates that the circuit topology is certainly feasible, but the model is incapable of fully replicating the gain adaptation Fulton posits in his theory.
The analogue processing involves several stages of adaptation, distribution and feature extraction engines, before the analogue signals become the pulses at the ganglion cells. The pulse density of signals on the audio nerve bundle carries the information from the feature extraction neural network.
Dr. Staples concluded the lecture by challenging the members of the audience to read the full works summarised in the presentation, and to join the discussion about how the ear really works.

A spirited Q&A session following the talk indicated the amount of interest in this subject.

More About Melbourne Section

AES - Audio Engineering Society