3
$\begingroup$

It is quite simple to understand the concept of lateral localization of sound. It depends upon the loudness and time (and wave phase) difference between 2 ears.

But how can we detect front-back localization (if someone is calling me from my front/back direction) and up-down localization (such as an aeroplane's sound is coming from upper place and a car's whistle is coming from lower-direction of the balcony I'm standing). How does that happen? Or doesn't localization happen at all under these conditions? Is it just an illusion just because we are already aware about the possible source (such as since it is aeroplane's sound then it might coming from upward place, or I can see my friend is talking by standing my front so the sound must coming from front... etc.)

I've searched wikipedia and other sites, but I found nothing that explained the matter sufficiently.

$\endgroup$
3
$\begingroup$
  • Localization along the azimuth (horizontal left-right axis) is mediated by various processes: 1) First, there is the head shadow effect, which means that sounds from the left reach the right ear (AD) in an attenuated state relative to the left ear (AS) due to the presence of bone and other tissues. This difference is picked up by the superior olive (SO). This mainly aids in the localization of high frequency sounds, as low frequencies simply tend to go around the head, rather than traverse it. As mentioned by others, there is a time delay between the two ears, referred to as the 2) inter-aural time delay (ITD). Given the speed of sound this adds up to delays in the microsecond range, yet the SO is still able to detect such minute differences to aid in sound localization. There is also a 3) Phase difference between sound waves reaching AD and AS when sound comes from the left or right are also picked up registered. This cue is important for low frequencies (Mather, 2006).

  • Localization in the vertical plane (elevation) is less well investigated. As far as I know it is the shape of the outer ear (pinna) that transforms the frequency characteristic of incoming sound. In terms of directional hearing it is thought that sounds coming from above lead to a different head transfer function than sounds from below. This means that localization can only be accomplished when the sound has a particular familiar characteristic that is slightly disturbed in the frequency domain when encountered at different angles of elevation (Hofman & Van Opstal, 2003).

  • Front-back localization is thought to be mediated by processes similarly to elevation cues. Spectral modifications work like filters and are derived from the reflections of sound caused by the shape and size of the head, neck, shoulders, torso, and especially, by the pinnae (Zhang & Hartmann, 2010).

References
- Hofman & Van Opstal, Exp Brain Res (2003); 148: 458–70
- Mather, Foundations of Perception, Psychology Press (2006)
- Zhang & Hartmann, Hear Res; 260(1-2): 30

$\endgroup$
3
$\begingroup$

The Wikipedia article is quite good.

In brief, as you state, the wave phase can be used only to localise sounds in the plane of the ears. To have an approximation of the position in the median plane (above-below/front-back), we use asymmetries that distort the sound differently at different frequencies.

In humans, it seems that these asymmetries and their processing are not genetically programmed but evolved by each listener.

This is different in owls, although different species have each a specific asymmetry. https://en.wikipedia.org/wiki/Sound_localization_in_owls

$\endgroup$
1
$\begingroup$

The geometry of the head + external and internal ear makes an angle-dependent spectral filter. For rich sound which timbre is known we can characterize the spectral filtering, and thus the location (probably as well for moving sound). Contrarily, it has been shown that pure sounds produce illusions of localisation (e.g. changing the frequency changes the perceived angle). Moreover, we have some a priori expectations, e.g. high pitch sound being more likely above.

$\endgroup$
  • $\begingroup$ thank you for a response. I'll take time to understand the process. $\endgroup$ – Always Confused Nov 17 '16 at 18:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.