Transcript Binaural Hearing

Binaural Hearing
Or now hear this!
Upcoming Talk: Isabelle Peretz
Musical & Non-musical Brains
Nov. 22 @ 12 noon + Lunch
Rm 2068B South Building
TLA 6: 2 Two Ear Hearing
• Purpose of TEH
– Spatial hearing and understanding
• Activity:
– Walk rapidly down a hallway while plugging one ear
– Halfway through hallway, switch to plugging the other
ear
• Switch order of plugging the two ears and repeat
• Write-up
– Does having a plugged ear change how you walk
down a hall? How did changing the plugged ear affect
your motion?
Hearing Binaurally
(Yost chapter 12)
• Binaural = two ear hearing
– Combination of information to determine spatial
position
• Azimuth
– Not distance
– Not vertical position
– Stationary localization
• Different cues available with motion
• Interaural cues for binaural hearing
– Interaural Loudness Difference (ILD)
• Interaural Intensity Difference (IID)
– Interaural Timing Difference (ITD)
– Interaural Phase Difference (IPD)
Interaural Timing Differences (ITD)
• Onset of auditory stimulation
– Does not vary across frequency
• Salient with lower frequencies (< 1500 Hz)
– Maximum delay of < 1 ms
• Dependent on head-size
• Angle of stimulation
• Critical for short events
– Clicks, bursts
• Less important for enduring events
– Noise, speech
Interaural Phase Differences (IPD)
• Relative phase of stimulus across ears
– Critical region is < 800 Hz
• No IPD at 833, 1666 Hz
– Noticeable differences of phase
• Minimum displacement 0.2 ms
• Enduring sound events
– Noise, speech
• Change in phase triggers change in localization
– Basis of the Precedence Effect
Interaural Loudness Differences
(ILD)
• Relative intensity across ears
– Critical region
• > 2 kHz
• Ecological constraints 800 Hz
– Up to 20 dB SPL attenuation (over 8 kHz)
• Sensitive to 1 dB SPL difference
• Total masking 8 – 10 dB SPL
– Similar to natural head shadow
• Oldest theory of directional hearing (1870’s)
• Ambulance direction
– Open window determines positions for high
frequency siren
Duality Theory of Directional
Hearing
• Frequency region determines salient cues
– Lower frequencies 40 – 1500 Hz IPD, ITD
– Higher frequencies 4 – 20 kHz ILD
• Worst localization performance 1500-4000 Hz
• Harnessing Stationary cues
– Difficult noises
• Diffuse noise, enduring
• Sinewave burst
– Easiest to localize
• Broadband click
– Incorporates multiple cues
Minimum Audible Angle (MAA)
• How good is hearing?
– Stationary: accuracy separating two
sound sources (Mills, 1958)
• Play sound, move left/right play again
• Chance performance = 50 %, threshold =
75%
– Results
• Azimuth dependence: best at center 0˚,
logarithmic decline to 75˚
• Frequency dependent: best 40 – 4000 Hz
– Approx. 3˚ separation (vision 1’)
• Minimum audible movement angle
– Velocity – dependent
• Approx. 1˚ separation
Localization with HAs
• Factors affecting localization
– Bilateral vs. Unilateral
• 2 ear vs. 1 ear
– Symmetric hearing loss?
• All sounds located at hearing ear
– If symmetrical bilateral improvement
• Speech in noise release from masking
– BTE vs. ITC/CIC
• BTE microphone outside ear canal
– Directional microphones
• ITE/CIC spectral filtering from pinnae
– Better HA performance with ITC/CIC
Localization with Cochlear Implant
• Test unilateral, bilateral cochlear implant
users
– ITD, IPD cues
– ILD cues
• HYPOTHESES?
• 3x precision with bilateral implants
– Large individual differences
• Duration using bilateral implants
• Speech ability
Head-related Transfer Functions
(HRTFs)
• HRTF: calculation of the sum of spatial parameters
– Distance between the ears
– Pinna filtering
• Spectral shape of resonance harmonics
– Head attenuation
• Nose directionality
• Body absorption
• Hair on the head
• Calculation of HRTF for simulated reality
– Convolve microphone input
– Dummy-head recordings
– Binaural recordings
• Which is best?
• Front-back confusions
Binaural Masking
• Vary position of noise & energetic masker
• Monaural
– No difference of spatial position and noise
– Similar amount of energetic masking in all positions
• Diotic
– No difference of spatial position of noise
– Similar amount of energetic masking
• Dichotic
– Noise to one ear, masker to other
– Release from masking
• Better detection of signal
Hearing the Silent World
• Localization
– Study of sound sources
• Sound producing objects relative to listener
• Are sound sources the basis of hearing?
– Visual world
• Light producing objects
– Sun, lamps
• Light reflecting surfaces
– Tables, faces, trees
– Can we detect sound obscuring/reflecting surfaces?
Hearing the Silent World
• Sound obstructing surfaces
– Diffuse sound field set behind sound
attenuating surfaces
• Are listeners sensitive to position of
surfaces?
• Test behavioral judgment
– Is the aperture large enough to allow
passage?
• Ego-centric judgment facilitates
accuracy
– Aperture size affects intensity,
spectra
• Randomize intensities, sine wave
signals
– Listeners can detect position of
sound obstructing surfaces
Elevation
• Height relative to listener
– How can this be determined?
• Interaural cues?
– Timing difference between the ears
• Mid-Saggital plane
– Loudness difference between the ears
• Absorption by head & pinna
– Front-back confusions
• Pinna cues
– Forward, downward facing
– Partially resolve front-back errors
Distance
• How far away is a sound
source?
– Interaural cues?
• Azimuth does not indicate
relative distance
– Pinna cues?
• Slight-downward facing
– More distant cues higher
in the perceptual plane
• Salient cues for distance
– Intensity
• Attenuation over distance
– Frequency dependent
• Unreliable indicator
– Reverberation
• Increase in number and
lag of echoes
– DEMO
Improving Accuracy
• How do listeners judge distance?
– Metrics of perception
• Absolute distance: objective scale
• Egocentric distance: metric in body relations
• Test
– Judge baby rattle distance egocentric scale
• 1 vs. 2 degrees of freedom
– Arm vs. Arm + body lean
– Highly accurate judging 1 or 2 degrees
• Better accuracy than found with absolute distance