Introduction to Psychoacoustics

1
Introduction to Psychoacoustics

• By Carolina Figueroa
• University of Idaho
• Advanced Human Factors

2

• Spatial sound is 3D (like vision)
• Azimuth cues left to right
• difference in the times at which sound waves
  arrives at the 2 ears
• Elevation cues up or down
• spectral changes produced by outer ear (or
  pinnate)
• Distance cues near or far
• Applications
• computer games,
• aids for vision impaired,
• virtual reality systems,
• eyes-free displays for pilots and air-traffic
  controllers,
• spatial audio for teleconferencing and shared
  electronic workspace

3

• Generating 3D sounds
• 3D sound loudspeakers at many different
  positions (expensive).
• 2 ears 2 channels, it is possible generate 3D
  with 2 channels (binaural approach using HRTFs)
• HRTF Head-Related Transfer Functions
• Function of the location of the source relative
  to listener
• Captures spectral changes caused by torso, head,
  outer ears (physical size and shape of listener)
  when a sound wave propagates from sound source to
  listeners ears. These changes depend on the
  azimuth, elevation and range (distance) from
  listener to source.
• Sound signal filtered by accurate HRTF and sent
  to the 2 ears of listener (headphones) is
  experienced as 3D sound

4

• Major factors that influence spatial hearing
• Coordinate system
• Azimuth cues
• Elevation cues
• Range cues
• Reverberation and Echoes

5

• Coordinate systems

XY horizontal XZ frontal YZ median
Spherical coordinates Azimuth angle over from
median plane (right, left) Elevation angle from
horizontal plane to source and X axis
(up,down) Range (near, far)
6
Duplex Theory. There are two primary cues for
azimuth Interaural Time Difference (ITD) and
Interaural Level Difference (ILD). ITD a/c (?
sin ?) /- 90 deg ITD max (source off to one
side) a/c (3.14/2 1) ITD 0 sound is directly
ahead a distance source c sound speed 343
m/s ILD (frequency dependent) ILD low (1.5 kHz)
almost no difference in sound presure at 2
ears ILD highv (gt 1.5 kHz) 20dB or greater
difference head-shadow effect

• Azimuth cues (binaural cues)

0.7 ms difference between time to arrive to right
ear and left for sound in the horizontal plane
7
Azimuth
Altitude (elevation)
8
Outer ear or pinna acts acoustic antenna Resonant
cavities amplify some frequencies and geometry
attenuate others Its response is directionally
dependent 2 paths from source to ear canal (a
direct and longer path) Low freq. collects
additional sound energy and 2 paths arrive in
phase High freq. long path is out of phase with
the direct signal and destructive interference
occurs Greatest interferance path length is
half wavelength producing a pinna
notch Pronounce pinna notch for sources
ABOVE Path differences changes with elevation,
pinna notch moves with elevation

• Elevation cues (monoaural)

Pinna shape varies from person to person causing
shifts in frequencies, harder control elevation
(individual HRFTs)
9

• Range cues
• Humans are best estimating azimuth, next
  elevation and worst estimating range.
• Loudness
• Sound energy coming from source falls inversely
  with the square of range.
• No 1-to-1 relationship between emitted and
  received sound energy
• Playing a soft sound does not mean is far
• Need to know characteristics of source
• Motion parallax
• Translation of head, the change azimuth depended
  on range.
• Close sources, small shift, large change in
  azimuth distant sources, almost no shift, no
  azimuth change
• Excess interaural level difference
• ILD increases when very close sources to the head
• Extreme case insect buzzing in one ear
• Ratio of direct to reverberant sound (major cue)
• Ordinary rooms sound is reflected and scattered
  from environmental surfaces
• At close ranges ratio is large, long rages ratio
  is small

10

• Reverberation and Echoes
• Reflections time delay 30-50 ms echoes
• Anechoic chambers absorb sound energy, only
  radiated energy reaches the ears
• Reflections do not interfere with ability to
  localize sources because we adapt quickly

11

• Spatial audio systems
• 2-channel (stereo)
• place a sound on the left, send its signal to
  the left loudspeaker, to place it on the right,
  send its signal to the right loudspeaker
• Multi-channel (surround)
• separate channels for every desired direction,
  including above and below Binaural recordings
• Expensive and unlikely to play role in HCI

12

• Binaural recordings
• recreate the sound pressures at the right and
  left ear drums that would exist if the listener
  were actually present
• Several disadvantages
• use of headphones
• not interactive, but must be prerecorded
• If the listener moves, so do the sounds
• Sources that are directly in front usually seem
  to be much too close
• pinna shapes differ from person to person,
  elevation effects are not reliable
• Improvements using HRTFs

13

• Head-Related Transfer Functions (HRTF)
• Fourier transform of the head-related impulse
  response sound pressure from the source to the
  ear drum
• HRTF captures all physical clues to source
  localization
• Once HRTF for right and once for left ears
• HRTF 4 variables 3 space coordinates and
  frequency
• Most HRTF measurements are made in far field,
  HRTF falls inversely with range
• Far field for sources at distances greater than
  1 meter because reduces HRTF to a function of
  azimuth, elevation and frequency
• HRTF are measured in anechoic settings (dont
  include effects of environmental sound
  reflections)
• Binaural room simulator to introduce important
  reflections so when using headphones you wont
  hear sounds very close to or inside or the head

14

• HRTF-Based systems
• Able to produce elevation, range and azimuth
  effects
• Person-to-person differences and computational
  limitations, it is much easier to control azimuth
  than elevation or range
• Headphones uncomfortable, have to be compensated
  from resemble pinna responses, compensation is
  sensitive to position.
• Loudspeakers problems with low frequencies,
  distance between speakers
• Head tracking recalculate relative position
  (location and orientation of the head) of each
  source modifying HRIRs too, is expensive, not
  very reliable, latency, unwanted transients
• Measured vs. Modeled HRTFs standards, set of
  standards (group of people), individualized or
  use a model HRTF (rational function, series
  expansions, structural models)

15

• Facts
• Best sources 90deg in horizontal plane, directed
  to right ear
• Weakest source at 270 deg or -90 deg on the
  opposite side of the head
• Front/back (0 and 180 deg) are similar