LE 12 Localization of Sound
Localization of Sound
Challenges in Auditory Localization
The auditory system struggles with sound localization due to the inability to preserve spatial information as seen in vision.
In vision, light reflects off objects in straight lines preserving object relationships on the retina; in contrast, sounds combine into a single waveform received by the ear.
Each sound source must be separated by the auditory system to differentiate between them (e.g., identifying a person speaking or a car horn).
Sound Positioning Concepts
Sounds can be thought about along two dimensions:
Azimuth: Position along the horizon.
Elevation: Vertical position of sound.
Each sound source is mapped onto an imaginary sphere surrounding the head.
Behavioral Data on Sound Localization
Participants estimate sound localization in a test arena with different sound positions.
Results show:
Greater accuracy for sounds coming from the front (blue dots closer to red squares).
Reduced accuracy for sounds from the side or behind.
Individuals can orient towards sounds for better resolution.
Cues for Sound Localization
Main Cues
There are three main cues utilized for sound localization:
Interaural Time Differences (ITD): The timing difference between sounds arriving at the two ears.
Interaural Level Differences (ILD): The intensity or loudness difference between sounds reaching each ear.
Head-Related Transfer Function (HRTF): The specific way sound modifies as it interacts with the head and ears.
Binaural Cues
ITD and ILD require two ears:
Interaural Time Difference: Depending on the sound's location, waves will reach one ear before the other.
The maximum ITD occurs for sounds located 90 degrees to the side, approximately 0.6 milliseconds for an average headwidth.
Interaural Level Difference: Sounds are affected by the head’s size and density; reflections and absorption alter sounds, resulting in varying levels of intensity entering each ear based on the sound's position.
Monaural Cues
HRTF provides information relevant to each ear, independent from the other.
Everyone’s ears have different HRTFs due to anatomical differences, which allow the brain to determine elevation.
Head-Related Transfer Function (HRTF)
Definition of HRTF
The HRTF is how sound changes in amplitude and frequency as it interacts with a person’s head and ears.
The HRTF varies between individuals and across different sound frequencies, leading to perceptual differences in sound quality.
Mechanics of HRTF
For instance, cupping hands behind ears alters incoming sounds by changing reflections and amplifications.
Manipulating the shape of the ear affects the frequency spectrum of sounds, leading to variations based on sound position.
Sound Localization Mechanisms
Interaural Time Differences (ITD)
ITDs inform on sound location via timing differences:
Absolute time differences are small but significant enough (e.g., from 0 to approximately 0.6 milliseconds).
As conditions change (left vs right), the ITD shifts systematically.
Interaural Level Differences (ILD)
Sounds are perceived to be louder from the side they are coming due to reflections & sound absorption by the head.
Higher frequency sounds are better at creating ILDs; low frequency sounds usually do not provide useful localization information.
The Cone of Confusion
Occurs when ITD and ILD cues yield similar values for different spatial locations, leading to difficulty in sound source localization.
Resolved by extracting timing information through individual frequency channels (phase locking).
The Role of the Nervous System
Neural Processing of Sound Localization
Auditory information is processed first at the brainstem, notably in the cochlear nucleus, where auditory nerve cell bodies reside.
The signal then projects to the superior olive, where ITDs and ILDs are compared via circuits, including coincidence detection mechanisms called Jeffress coincidence detectors.
Jeffress Coincidence Detector
A proposed mechanism for processing ITDs; sensitive neurons located in neural arrays respond to phase-locked action potentials from sounds reaching each ear.
Auditory Scene Analysis
Auditory Scene and Analysis
Defined as the array of sound sources in the environment and their subsequent analysis into separate perceptions.
The problem of auditory scene analysis involves distinguishing between overlapping sounds (e.g. conversations at a party).
This relates to the cocktail party problem where one must detect and allocate focus on specific sounds amidst a cacophony.
Cues for Auditory Scene Analysis
Sounds can be grouped based on:
Proximity in space (sounds from similar locations).
Proximity in time (sounds occurring in rapid succession likely from the same source).
Similarity in perceptual qualities (pitch, timbre).
Examples in Music and Environment
Auditory perception, such as in music, utilizes similarities in pitch and proximity to create melodies perceived as separate streams despite being played on the same instrument.
Composers like Bach utilized principles of auditory scene analysis to create distinct musical lines.
Integration of Auditory and Visual Information
Ventriloquism Effect
Highlights how visual cues can dominate auditory ones, leading to the perception that a sound originates from a seen object rather than its true source.
In experiments, auditory and visual signals can influence the perception of events (e.g., perceiving one vs. two flashes based on auditory cues).
Environmental Context
The auditory system's integration with visual information aids in creating a coherent representation of complex auditory scenes.
Conclusion and Exam Preparation
Key Takeaways
Localization of sound involves complex auditory processing using binaural cues (ITD, ILD) and monaural cues (HRTF).
Understanding auditory scene analysis is crucial for effective communication in complex auditory environments.
Upcoming Review Session
An exam will be held in a week; a review session is scheduled for the Tuesday before the exam. Students encouraged to attend for clarification on study material.