How does sound localization vary with distance and frequency?

[04shaikj]

Hi there,

This is not a homework question. I am currently conducting a visual perception experiment with a professor. We are trying to find out how duration of a sound presented concurrent with a visual stimulus affects the localization of the visual stimulus which participants have to localize. The experiment protocol is as follows.
1. There is a a fixation point on the screen that subjects stare at.
2. They click a mouse and a target appears to the left or right of the fixation point (which dissappears) and the participant has to make a saccade (eye movement) to that target.
3. Around the same time a stimulus is briefly flashed on the screen concurrent with a sound (white noise) played for some duration (8 ms).
4. Subjects have to use the cursor to click on where they think they saw the flash occur.

The issue is with step # 3. We want to make sure that the sound is perceived to emanate at the location where the flash occurs and not just coming out of both speakers. Using stereo doesn't work because that just plays the sound from either the right or the left speaker. Here's to show you what I mean:

FP--------Flash
1
1
1
Head

FP = Fixation point on the screen.
Participant's head always faces straight (we have a bite bar to keep their head stable)
The visual stimulus is flashed some random degrees to the right of left of the FP (above it's flashed to the right). I am looking for a function that shows amplitude (loudness) decreasing as a function of increasing distance (between head and flash). I don't where to start since I only have a introductory background in Physics.

 

[atyy]

Something like this?
Krukowski AE, Stone LS.
Expansion of direction space around the cardinal axes revealed by smooth pursuit eye movements.
Neuron. 2005 Jan 20;45(2):315-23.

You may have to vary the time-delay between the speakers.

You may also have to consider the head-related transfer function of each participant.

A standard reference is Blauert's Spatial Hearing.
http://books.google.com/books?id=wB...&hl=en&sa=X&oi=book_result&resnum=4&ct=result

 

[Dale]

Hi 04shaikj, welcome to PF

Atyy already hit on the key part of this. One of the main cues for localizing sound is time delay (phase shift). I don't know if white noise can be localized, clicks might work better.

 

[04shaikj]

Hey thanks for the replies and book reference. I will definitely check that out. I can't say I totally understand what you are saying because I don't see how time delays between speakers gives the perception of sound emanating from exactly where the flash occurred.
For example, above where the flash occurs to the right of the fixation point, you would 1st want to know the intensity of the sound when flashed at the FP. Then subtract some function from that the further the flash (and thus, the sound) occurs away from the FP so that the intensity of the sound decreases. So essentially in the above example where the flash occurs to the right, I would use stereo to make the sound emanate from the right speaker and vary its intensity depending on where the flash occurs...

 

[atyy]

I can't say I totally understand what you are saying because I don't see how time delays between speakers gives the perception of sound emanating from exactly where the flash occurred.

Suppose the flash actually emitted a sound. The sound will (1) take different amounts of time to travel to each ear depending on the distance travelled, (2) and be attenuated in intensity by different amounts at each ear depending on which parts of the head was between the ear and the sound. The brain uses both interaural time difference and interaural level difference to estimate spatial location.

Low frequencies have longer wavelengths, so they are not so easily attenuated by the head, so they are usually localized using interaural time difference. High frequencies are more easily attenuated, so they are usually localized using interaural level difference.