Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.623697
Title: Aural discrimination of pitch in the presence of noise
Author: White, Richard E.
Awarding Body: University of London
Current Institution: Imperial College London
Date of Award: 1970
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis describes an investigation of the effects of additive noise on the discrimination, by human listeners, of the relative pitch of two sinusoidal sounds. The experiments were designed to test an Ideal Pitch Discriminator model, an important implication of which was that the ear could match itself to a signal in a noisy situation. However, the results were not consistent with this model, and an explanation in terms of the Critical Band effect was preferred. This effect had not previously been demonstrated in a pitch discrimination task. The Difference Limen for pitch varied with SN ratio according to a curve of characteristic shape, located on the SN axis at a position dependent on the noise band-width. The separation between wide and narrow band curves was used to estimate the effective bandwidth of the ear. A lumped-constant filter model was constructed to simulate the frequency dependent behaviour of the middle ear and basilar membrane. This model could account for the frequency selective (or Critical Band) properties inferred from the subjective experiments. It was also proposed that the perception of pitch was based on the temporal properties of auditory nerve impulses initiated by the zero-crossings of oscillatory basilar membrane movement. It was argued that the generation of such impulses was analogous to the zero-crossing detection performed by an ideal FM discriminator, and it was shown that the effects of noise on such a device could account for the subjective data. A quite separate aspect of this thesis was the evaluation of a Sequential testing and estimation method. Although a Sequential strategy appeared to cause some instability in a subject's responses, it had a number of advantages over the more commonly used Probit Analysis, particularly for the estimation of Difference Limens.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.623697  DOI: Not available
Share: