Jump to content

Talk:Head-related transfer function

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Untitled

[edit]

"Humans are less sensitive to changes in the azimuth, φ"

We are MORE sensitive to changes in azimuth than changes in elevation, aren't we ? --Adoudoux (talk) 13:11, 8 February 2008 (UTC)[reply]

This entry, http://wiki.riteme.site/wiki/Sound_localization, suggests so, "... Our eardrums are only sensitive to the sound pressure level differences.

Note that these cues will only aid in localizing the sound source's azimuth (the angle between the source and the sagittal plane), not its elevation (the angle between the source and the horizontal plane through both ears)..." —Preceding unsigned comment added by 209.137.158.52 (talk) 00:17, 15 April 2008 (UTC)[reply]

Alternative introduction

[edit]

In response to the request for a more understandable introduction to the subject I suggest the followig:

start intro

A head-related transfer function (HRTF) defines the response that an ear receives from a sound from a particular point in a three dimensional space. When sound from an external source meets a head, it "flows" around the shape of the head and reflects off various features like the external ears (pinnae). The ear canal receives two versions of this sound, modified by the shape of the head and the spacing of the ears, and the inner ear sends responses to the auditory cortex of the brain.

The shape of the head and pinnae form a complex diffraction pattern according to the direction of the sound source. By a heuristic process, the brain has learnt to recognise the direction of the sound source, and also to discriminate one source out of many in a complex noise field. The responses from two ears, each with their respective HRTF, are interpreted by the auditory cortex to give directional information from each distinct sound source and from general less defined sound sources. This is the mechanism behind the cocktail party effect.

end intro

keoka 02:01, 12 July 2012 (UTC) — Preceding unsigned comment added by Keoka (talkcontribs)


examples

[edit]

Would be nice to have some examples (algorithm wise) - or at least point out some usable implementations. draeath (talk) 22:58, 6 March 2012 (UTC)[reply]

Laplace rather than Fourier transform

[edit]

The statement "The HRTF is the Fourier transform of HRIR" is not quite correct and ought to be changed to "The HRTF is the Laplace transform of HRIR". The Fourier transform of HRIR is the frequency response, which is the transfer function evaluated on the s-axis. Sigmundg (talk) 08:03, 4 June 2024 (UTC)[reply]