Applied Psychoacoustics Lecture 12: Room Acoustics

1
Applied PsychoacousticsLecture 12 Room
Acoustics Perception in Virtual Auditory
Environment

• Jonas Braasch

2
W.C. Sabines Psychoacoustics

• Sabine started to measure the reverberation time
  with a stop watch by ear
• 60 dB was his dynamic range of hearing
• (from Manfred Schroeder, 1978)

3
Monaural Room Parameters

• Reverberation
• Initial Time Delay Gap
• Definition
• Clarity
• Center Time

4
Binaural Room Parameters

• Incoherence
• Apparent Source Width
• Listener Envelopment

5
Important Psychoacoustic Effects

• Post masking
• Precedence Effect
• Localization
• Interaural cross correlation

6
Cases were the rating for The pit sound was
statistically higher
at least 6 judgements per data point
Beranek, 2004
7
(No Transcript)
8
Impulse Response, Audimax Bochum
Braasch, 1999
9
Reverberation time limen

• Plenge (1965) measures reverberation time
  difference limens for various frequency bands (2
  bark band wide)
• Gun shot was the source signal
• All curves were monophonic
• Sound was sent through frequency filter
• one filter was equipped with time-dependent
  exponential envelope filter triggered by signal
  onset (oscilloscope)

10
Reverberation time limens
Reduced RT in freq. band
Enhanced RT in freq. band
Signal I
Signal III
Plenge (1965)
11
Reverberation time limens
Reduced RT in freq. band
Enhanced RT in freq. band
Plenge (1965)
12
Temporary masking
Forward Masking will partially mask the early
reflections
13
Seraphims Experiment (1961)

• Seraphim measured the masked detection threshold
• One lead (70 dB), one lag (variable level)
• Used speech signal
• Open circles 2AFC
• Closed circles method of adjustment

rel. Sound pressure level dB
delay
14
Measurement of the psychophysical function hf(s)
15
2AFC vs. method of adjustment
Method of adjustment
2 alternative forced-choice (AFC)
1
2
time
time

• Adjust level of reflection until it is just
  audible (detection threshold)

• Indicate which of the two stimuli contains a
  reflection (forced choice).
• One stimuli always contains a reflection (at
  different levels) the other does not
• Chance of guessing 50

16
A naïve assumption
100
detection threshold
Number of correct responses
0
Sound pressure level of reflection
17
it rather looks like this
75 threshold
Probability for correct response
50 threshold
Log-normalized stimulus intensity (e.g, sound
pressure level)
In signal detection theory, we explain this
variation with internal noise in the central
nervous system
18
Definition of the correct response
Positive response
Negative response
Stimulus present
Stimulus not present
Sometimes it is better to rather accept a false
alarm (e.g., fire detector) while other times it
is better to accept a miss (e.g., non-emergency
surgery cases)
19
Results (Constant Method)
20
Determination whether difference is perceivable
Internal response to stimulus 1, e.g.,
perceived loudness
Internal response to stimulus 2
21
(No Transcript)
22
(No Transcript)
23
Discriminability index dms-mm/(s2), if sssm
24
Receiver Operating Characteristic (ROC)
25
Seraphims Experiment (1961)
rel. Sound pressure level dB
rel. Sound pressure level dB
delay
delay
26
Seraphims Experiment (1961)
O direct sound source I fixed reflection T
test reflection
after Marshall and Barron, 2001
27
Seraphims Experimental Set-Up
28
Reichardt Schmidt 1967
RG noise generator MS tape recorder KM
microphone PS total level VM tape delay PO
direct level PW wall reflection level PD
ceiling level PN reverb level TP/HP low pass
/high pass VE reverb generator LV
amplifier PM microphone AM level meter a
azimuth b elevation L1, L2 direct sound L3
wall reflection L4 ceiling reflection L5-L8
reverb
29
Reichardt Schmidt 1967
change
Expected results
impulse response
judgement
change
30
Reichardt Schmidt 1967
level decrease
level increase
Absolute threshold
Measured (significance region) estimated
Output level
31
Reichardt Schmidt 1967
reverb reverb
Judgment of correct responses
Initial wall reflection level
Wall reflection level above direct sound level
l
32
Difference Limens for refl. level

• Reichardt and Schmidt determined the Difference
  Limen (just noticable difference) for the level
  of the first lateral reflection.

plotted after Barron Marshall, 1981
33
Difference limens for refl. level

• Reichardt and Schmidt determined the Difference
  Limen (just noticable difference) for the level
  of the first lateral reflection.

plotted after Barron Marshall, 1981
34
Barron Marshalls Set-up
Barron Marshall, 1981
35
Barron Marshalls Set-up

• Detection Threshold for a reflection
• in presence of the direct sound (dots)
• in presence of the direct sound and one ceiling
  reflection (x)

Barron Marshall, 1981
36
Levels of equal Spatial Impression
Barron Marshall, 1981

• Levels of equal Spatial Impression for a
  reflection pair with different delay times
  compared to a reflection pair at 40 ms (dots)
• Reference reflection is varied
• xs show the same data for a single reflection
• Music Sample (Mozart)
• The data shows the mean and 95 confidence
  interval

37
Levels of equal Spatial Impression II

• Same as last figure but for Wagner motif (dashed
  line shows Mozart motif)

Barron Marshall, 1981
38
Levels of equal Spatial Impression III

• This time the spatial location of the reflection
  pair is varied instead of the delay time

Barron Marshall, 1981
39
Schroeder et al. (1974)

• Correlation between listener judgments and
  objective parameters for 22 Concert halls.
• 12 listeners
• 22 European Concert Halls
• Two-track dummy-head recording
• Mozart Jupiter Symphony
• Reproduction in anechoic space with loudspeakers
  (cross-talk cancelation filters)
• Metric linear factor analysis

40
Loudspeaker reproduction
Schroeder et al. (1974)
41
Results

• Treverberationtime
• VVolume
• WWidth
• GITDG
• DDefinition
• CCoherence

11 Halls RTlt2.2sec
Schroeder et al. (1974)
42
Results

• Treverberationtime
• VVolume
• WWidth
• GITDG
• DDefinition
• CCoherence

11 Halls RT 2.0-3.2 sec
Schroeder et al. (1974)
43
Apparent source width

• Apparent source width (ASW) is the apparent
  auditory width of the sound field created by a
  performing entity as perceived by a listener in
  the audience area of a concert hall.
• It is generally accepted ASW can be determined
  from the sound field reaching the ears of a
  listener in a concert hall or opera house in the
  first 80 ms after (and including) the arrival of
  the direct sound from the source on stage. It is
  measured in degrees.

44
Listener Envelopment (LEV)

• Listener envelopment (LEV) is the subjective
  impression by a listener that (s)he is enveloped
  by the sound field, a condition that is primarily
  related to the reverberant sound field. The
  reverberant sound field is generally said to
  begin 80 ms after arrival of the direct sound.

45
Timbre
Timbre is that attribute of auditory sensation in
terms of which a listener can judge that two
sounds similarly presented and having the same
loudness and pitch are dissimilar.
ANSI, 1960
46
Relative Level, Strength Index, or Sound Strength
Free
47
Lateral Energy Fraction
48
Lateral Energy Fraction
49
Lateral Energy
50
Kuhl 1978
Linear fit to data for four concert halls for
various dynamic settings pp, mf, ff and settings
of Strength G, and Lateral Reflection LF
51
Running Liveness (Reverberant to Early Sound
Ratio)
52
Definition
53
Clarity (C50 or C80)
54
Center Time (centroid)
55
Tonal Color, Timbre
56
Tonal Color, Timbre
57
Temporal Diffusion
58
Interaural cross correlation
59
Interaural cross correlation
Interaural Incoherence 1-k
60
Beranek, 2004
61
Early Decay Time
Beranek, 2004
62
Beranek, 2004
63
Beranek, 2004
64
Beranek, 2004
65
Beranek, 2004
66
Beranek, 2004
67
Beranek, 2004
68
Beranek, 2004