Spectral cues to sound source localization

  • Toledo, Daniela (Project Participant)
  • Møller, Henrik (Project Participant)

Project Details

Description

Head-related transfer functions (HRTFs) contain amplitude and phase information that is used in sound source localization.  These transfer functions can be separated into temporal and spectral components, both of which are relevant in localization. Spectral components are highly dependent on anthropometric characteristics of subjects, particularly in the shape of their pinnae.  The individual differences can be contained in narrow frequency bands -e.g. a shift in center frequency of a dip or peak, a different bandwidth or Q factor- or in broad frequency bands -e.g. the absence of a dip or peak.  Even though the importance of the individual spectral characteristics as localization cues is widely accepted, it remains unclear whether relevant cues relate to narrow or broad frequency bands.  This is an important issue in the context of binaural synthesis with non-individual HRTFs: sound localization is degraded if the spectral characteristics of the HRTF used do not match well that of the listener.  It is still not well understood which spectral characteristics are to be matched and how. 

The PhD project that is being conducted investigates the spectral characteristics of HRTFs that are more relevant as localization cues and how to characterize them.  The hypothesis is that particular spectral features can be identified as localization cues -either in narrow or broad frequency bands- by cross-matching HRTFs from different subjects according to the results of localization experiments.  It is also hypothesized that these cues can be described by adequate parameters in time or frequency domain -e.g., arrival time, amplitude, center frequency, bandwidth, etc. 

StatusFinished
Effective start/end date19/05/201031/12/2017

Funding

  • <ingen navn>

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.
    • The role of spectral features in sound localization

      Toledo, D. & Møller, H., 2008, Audio Engineering Society Convention Papers: Papers 7314-7500. Audio Engineering Society, 6 p.

      Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

      File
      3 Citations (Scopus)
      53 Downloads (Pure)