"From
ITU-T Workshop on
Speech to Audio: bandwidth extension,
binaural perception"
Lannion, France, 10-12 September 2008
Loudness: Current Knowledge and
Questions
Sabine Meunier,
Laboratoire de Mécanique et
d’Acoustique – CNRS - France
Lannion, France, 10-12 September 2008
International
Telecommunication
Union
Loudness
Supraliminary sensation: how to measure it ?
Weber-Fechner law, 19th century
Weber: ΔΦ/Φ=constant
Φ:stimulation
Fechner:ΔΨ=k ΔΦ/Φ
Ψ: sensation
Ψ = A LogΦ + B
Stevens law: direct measurement,
magnitude estimation
Ψ = a Φb
Lannion, France, 10-12 September 2008
International
Telecommunication
Union
2
Methods to measure Loudness
ƒ Magnitude estimation: loudness in sones
1 sone = loudness of a 1-kHz tone at 40 dB SPL
ƒ Adjustement (loudness matches): loudness level in phons
a loudness level of a sound of X phons means that the sound is as
loud as a 1-kHz tone at X dB SPL
ƒ Adaptive (2down–1up, 1down–2up): loudness level in phons
ƒ Multitracking: loudness level in phons
ƒ
Categorical loudness scaling
International
Telecommunication
Union
Lannion, France, 10-12 September 2008
3
Methods to measure Loudness
Adjustment
ƒ Test sound (T): the sound that we want to know the
loudness
ƒ Sound of comparison (C): usually 1-kHz tone, variable
level
ƒC
after T (1st test) and T after C (2nd test)
ƒ Listener’s
task: Adjust the C-level to have the same
loudness as T
ƒ Random order, different for each listener
ƒ Start level was randomly X dB above or under T loudness
level
ƒ Loudness
International
level in phons : mean of C-level obtained
in
Telecommunication
the 2 tests for each sound
Lannion, France, 10-12 September 2008
Union
4
Methods to measure Loudness
ƒ Magnitude estimation: loudness in sones
1 sone = loudness of a 1-kHz tone at 40 dB SPL
ƒ Adjustement (loudness matches): loudness level in phons
a loudness level of a sound of X phons means that the sound is as
loud as a 1-kHz tone at X dB SPL
ƒ Adaptive (adown–bup, bdown–aup): loudness level in phons
ƒ Multitracking: loudness level in phons
ƒ
Categorical loudness scaling
International
Telecommunication
Union
Lannion, France, 10-12 September 2008
5
Methods to measure Loudness
adaptive methods
Presentation order : T & C or C & T, randomly
Listener’s task : Which of these 2 sounds is louder
2down-1up (1st test) :
Start level above the T loudness level
++
2nd reversal
-5 dB
++
Mean of the last Y reversals
++
-5 dB
-2 dB
+5 dB
-
-
+ JL
1st reversal
Track finishes after X reversals
2up-1down (2nd test) :
1st reversal
+5 dB
--
+
+
-5 dB
+5 dB
--
--
+2 dB
2nd reversal
Start level under the T loudness level
Lannion, France, 10-12 September 2008
Loudness
level in phons
- JS
International
Telecommunication
Union
6
Methods to measure Loudness
ƒ Magnitude estimation: loudness in sones
1 sone = loudness of a 1-kHz tone at 40 dB SPL
ƒ Adjustement (loudness matches): loudness level in phons
a loudness level of a sound of X phons means that the sound is as
loud as a 1-kHz tone at X dB SPL
ƒ Adaptive (2down–1up, 1down–2up): loudness level in phons
ƒ Multitracking: loudness level in phons
ƒ
Categorical loudness scaling
International
Telecommunication
Union
Lannion, France, 10-12 September 2008
7
Methods to measure Loudness
Multitracking
ƒ Principle similar to the adaptive method
ƒ 4 or 5 simultaneous sequences
ƒ Random choice of the sequence, different for each listener
ƒ Loudness level in phons:
mean of +JL and –JS for each sound
International
Telecommunication
Union
Lannion, France, 10-12 September 2008
8
Methods to measure Loudness
ƒ Magnitude estimation: loudness in sones
1 sone = loudness of a 1-kHz tone at 40 dB SPL
ƒ Adjustement (loudness matches): loudness level in phons
a loudness level of a sound of X phons means that the sound is as
loud as a 1-kHz tone at X dB SPL
ƒ Adaptive (2down–1up, 1down–2up): loudness level in phons
ƒ Multitracking: loudness level in phons
ƒ
Categorical loudness scaling
International
Telecommunication
Union
Lannion, France, 10-12 September 2008
9
Methods to measure Loudness
Categorical loudness scaling
50
too loud
45
very loud
40
35
Loudness measured in
Categorical Unit (CU)
loud
30
25
medium
20
15
soft
10
5
very soft
0
inaudible
Lannion, France, 10-12 September 2008
International
Telecommunication
10
Union
Loudness as a function of
SPL
N=k(P-P0)a
at 1 kHz: a=0.6
20
10
5.0
N: loudness
P: pressure
P0: constant
8000 Hz
4000 Hz
2.0
1.0
0.5
1000 Hz
0.2
0.1
.05
100 Hz
50
0H
z
Loudness in sones
200
100
50
.02
0
20
250 Hz
40
60
80
100
120
Sound pressure level (dB)
Lannion, France, 10-12 September 2008
International
From Scharf (1978)
in
Telecommunication
Handbook of perception,
11
Union
Carterette and Friedman
Loudness as a function of SPL
Partial loudness
50
Loudness in sones
20
10
5.0
2.0
1.0
0.5
0.2
Quiet 50
0.1
20
30
60
40
70
50
80
90
60 70
100 dB SPL
of Noise
80
90 100 110
SPL of masked tone (dB)
Lannion, France, 10-12 September 2008
From Scharf International
(1978) in
Telecommunication
Handbook of perception,
12
Union
Carterette and Friedman
Loudness as a function of
frequency
sound pressure level (dB)
Equal loudness contours
Standard: ISO 226, 2003
frequency (Hz)
Lannion, France, 10-12 September 2008
International
Telecommunication
13
Union
Loudness as a function of
frequency
Equal loudness contours
Sound Pressure Level (dB)
90
ISO226 2003
ISO226 1987
75 phons
80
70
60 phons
60
50 phons
50
From Boullet (2005)
40
100
Lannion, France, 10-12 September 2008
1000
Frequency (Hz)
10000
PhD Thesis
International
Telecommunication
14
Union
Loudness as a function of
bandwidth
Loudness level in phons
Spectral loudness summation
overall sound
presure level
critical band
Bandwidth (Hz)
Lannion, France, 10-12 September 2008
From Scharf (1978) in
Handbook of perception,
Carterette and Friedman
International
Telecommunication
15
Union
Loudness as a function of
duration
Temporal loudness summation
Loudness level (phons)
72
70
1-kHz pure tone
68
66
64
62
60
y = 71.2+10*log(1-exp(-x/0,127))
R² = 0,93
τ = 127 ms
Critical duration = 381 ms
58
56
0,01
0,1
Duration (s)
Lannion, France, 10-12 September 2008
1
From Boullet (2005)
PhD Thesis
International
Telecommunication
16
Union
Loudness models
Standards for steady sounds
ISO 532B, « Method for calculating loudness level »,
International Organisation for standardization (1975).
From Zwicker E., Acustica, 10, 304 (1960)
Zwicker E., J. Acoust. Soc. Am., 33, 248 (1961)
ANSI, S3.4-2005, « Procedure for the Computation of
Loudness of Steady Sounds, », American National
Standards Institute, New York (2005).
From Moore B. C. J. and Glasberg B. R., Acustica-Acta Acustica,
82, 335 (1996).
Moore B. C. J., Glasberg B. R., Baer T., J. Audio Eng. Soc.,
45, 224 (1997).
International
Telecommunication
17
Union
Lannion, France, 10-12 September 2008
Zwicker’s model
Stationary sounds, free or diffuse field
Signal
FFT
a0 filtering
From free field to inner ear
aD+ filtering
From free field to diffuse field
Critical bands
or 1/3 octave bands
1
Excitation
24
Calculated using masking curves
Specific loudness
Based on Stevens law
Overall loudness
=Σ specific loudnesses
Lannion, France, 10-12 September 2008
(because of spectral loudness
International
Telecommunication
summation)
18
Union
Moore and Glasberg’s model
Stationary sounds, free or diffuse field
Based on Zwicker’s model
Differences:
1 –Auditory filters shapes,
2 –Excitation pattern,
3 - a0 and aD+
International
Telecommunication
19
Union
Lannion, France, 10-12 September 2008
Loudness models for nonstationary sounds
ƒ Zwicker E., “Procedure for calculating loudness of temporally
variable sounds”, J. Acoust. Soc. Am., vol.62, n°3, 675-682, 1977.
ƒ Zwicker E. et Fastl H., “Psychoacoustics: Facts and models”, 2nd
Edit ion, Springer-Verlag, Berlin, 1999.
ƒ Glasberg B. R. and Moore B. C. J., “ A model of loudness applicable
to time-varying sounds”, J. Audio Eng. Soc., 50, n°5, 331-342,
2002.
International
Telecommunication
20
Union
Lannion, France, 10-12 September 2008
Current researches
ƒ Loudness of non-stationary sounds
Short duration signals
Long duration signals
ƒ Effect of context
Induce Loudness Reduction (Recalibration)
Loudness Constancy
Binaural Loudness Summation
ƒ Spectral loudness summation and duration
International
Telecommunication
21
Union
Lannion, France, 10-12 September 2008
Loudness of non-stationary sounds
Short duration sounds
Temporal integration =
Lshort – Llong
short and long signal at
equal loudness
15
Level Difference (L
short
-L
long
) in dB
20
Temporal integration
10
depends on level
5
Temporal integration
maximum for moderate
levels
Short varied
Long varied
0
20
40
60
80
100
Level of Short Tone in dB SPL
Lannion, France, 10-12 September 2008
120
International
Telecommunication
22
Union
From Florentine et al., JASA 1996
Loudness of non-stationary sounds
Short duration sounds
Modified Power Function
200-ms Equal-Ratio Loudness
5-ms Equal-Ratio Loudness
20
15
Level Difference (L
short
-L
long
) in dB
Loudness functions are not linear
10
Lshort – Llong =14dB
5
Short varied
Lshort – Llong =19dB
Long varied
0
20
40
60
80
100
Level of Short Tone in dB SPL
Lannion, France, 10-12 September 2008
120
International
Telecommunication
23
Union
From Florentine et al., JASA 1996
Loudness of non-stationary sounds
Short duration sounds
These loudness functions show features similar to the mechanical
input/output measurement at the basilar membrane
Temporal integration of loudness does not depends on level
International
Telecommunication
24
Union
Lannion, France, 10-12 September 2008
Loudness of non-stationary sounds
Environmental short-duration
sounds
Environmental sounds:
Exponential envelope
Most studies:
Rectangular envelope
N = kEaTb
N: loudness
E: energy
T: sound duration
International
a, b: constants
Telecommunication
Union
Lannion, France, 10-12 September 2008
From Meunier et al., ICA 2001, Forum Acusticum 2002
25
Loudness of non-stationary sounds
Environmental short-duration
sounds
N= kEaTb
Determination of a and b
a: loudness functions for environmental short duration sounds
b: loudness as a function of duration
International
Telecommunication
26
Union
Lannion, France, 10-12 September 2008
Loudness of non-stationary sounds
Environmental short duration sounds
Loudness as a function of duration
Three relationships between loudness and duration were found in
different studies :
- Loudness is constant when Energy is constant:
Equal energy rule
for duration less than the critical duration
loudness = constant if
Energy = Intensity x duration = constant
- Loudness is constant when Energy
decreases as duration increases
- Loudness is constant when Energy
increases as duration increases
Lannion, France, 10-12 September 2008
International
Telecommunication
27
Union
Loudness of non-stationary sounds
Environmental short duration sounds
Loudness as a function of duration
T
Energy =
2
E = ∫ p((t) dt = constant
0
International
Telecommunication
28
Union
Lannion, France, 10-12 September 2008
From Meunier et Rabau., Acoustics 08
Loudness of non-stationary sounds
Environmental short duration sounds
Loudness as a function of duration
Estimated loudness
(normalized)
88 dB HL
73 dB HL
58 dB HL
60dB SL
45dB SL
1
0,1
bands of noise
F0=1 kHz
Δf= 80 Hz
1
10
100
Duration (ms)
1000
When Energy constant, Loudness
varies as a power function of
Signal Duration.
Exponent depends on the level
of the signal.
Softer signals:
Loudness constant when Energy
decreases as duration increases
Louder signals:
Loudness constant when Energy
constant as duration increases
International
Telecommunication
29
Union
Lannion, France, 10-12 September 2008
From Meunier et Rabau., Acoustics 08
Loudness of non-stationary sounds
Loudness Model for Impulsive
Sound (LMIS)
90
80
70
60
50
40
Measured Loudness Level
LMIS
30
20
son23
son22
son02
son05_L3
son03
son24
son09
son15
son01
son05_L2
son06
son04
son13
son05_L1
son19
son16
son20
son08
son12
son11
son17
son18
son07
son21
son14
son10
Loudness Level (phons)
100
Sound n°
International
Telecommunication
30
Union
Lannion, France, 10-12 September 2008
From Boullet et al. in preparation
Loudness of non-stationary sounds
Long duration sounds
How does listeners judge overall loudness of
time-varying sounds ?
Loudness N
N10
N5
?
1 min
Time (s)
International
Telecommunication
31
Union
Lannion, France, 10-12 September 2008
Loudness of non-stationary sounds
Long duration sounds
Kuwano and Namba (Psychol. Res., 1985) and
Fastl (5th Oldenburg Symp. Psych. Acoutics,
1991):
Sound events proeminent in level strongly
influence global loudness
Susini et al. (Acta Acustica, 2002):
Recency effect: related to the temporal position
of the highest contour peak
Global loudness: combination of highest levels,
of their temporal position and their duration of
International
emergence
Telecommunication
Union
Lannion, France, 10-12 September 2008
32
Loudness of non-stationary sounds
Temporal asymmetry
Loudness change of tones with linearly varying levels
Loudness change: asymmetric
Asymmetry depends on:
direction of change (increasing vs. decreasing)
range of levels (high vs. low).
International
Telecommunication
33
Union
Lannion, France, 10-12 September 2008
John G. Neuhoff , Nature, 1998, 395, 123-124
Loudness of non-stationary sounds
Temporal asymetry
International
Telecommunication
34
Union
Lannion, France, 10-12 September 2008
John G. Neuhoff , Nature, 1998, 395, 123-124
Loudness of non-stationary sounds
Temporal asymetry
Ratio of estim ations
(Louder divided by softer)
100
A) - Pure tones at 1 kH z
75->45 decreasing
45->75 increasing
60->75 increasing
75->60 decreasing
(17 subjects)
10
1
1
10
100
Sw eep duration in seconds
Canévet et al., Acta
Lannion, France, 10-12 September 2008
International
Telecommunication
Acustica,
Union 2003 35
Effect of context
Induced Loudness Reduction (ILR)
A preceding higher-level tone (inducer) reduces the
loudness of a lower-level tone (test tone)
A: 500-Hz tones relatively
low SPLs and 2500-Hz
tones high SPLs
B: reverse
Marks, J. Acoust.
Lannion, France, 10-12 September 2008
International
Telecommunication
36
Union Am., 1996
Soc.
Effect of context
Induced Loudness Reduction (ILR)
Amount of ILR depends on:
Tone levels
Frequency separation between inducer and test tone
Duration of inducer and test tone
Time separation between inducer and test tone
Individual differences
Review in Epstein, J. Acoust. Soc. Am.,
Lannion, France, 10-12 September 2008
International
Telecommunication
37
Union 2007
Online,
Effect of context
Loudness Constancy
Intensity changes at the ear
may be due to
• Source power changes
• Source distance
source
power
source
power
Loudness Constancy
Loudness contant for fixed
source power and variable
source distance
Zahorik and Wightman, Nature
Lannion, France, 10-12 September 2008
source
power
International
Telecommunication
38
Union
neuroscience,
2001
Effect of context
Binaural Loudness Summation
(BLS)
Binaural loudness = A x monaural loudness
A: from 1.3 to 2 depending on study
International
Telecommunication
39
Union
Lannion, France, 10-12 September 2008
From Marozeau et al. J. Acoust. Soc. Am., 2006
Effect of context
BLS as a function of stimulus and
listening conditions
Stimuli
• Monitored Live Voice (MLV)
spondees
• Recorded spondees
• Tones
Listening conditions
• Earphone
• Loudspeakers
BLS for tones or recorded spondees > BLS for MLV
BLS for earphones > BLS for Loudspeakers
BLS in laboratory conditions > BLS out of the laboratory
International
Telecommunication
40
Union
Lannion, France, 10-12 September 2008
From Florentine J. Acoust. Soc. Am., 2008
International
Telecommunication
41
Union
Lannion, France, 10-12 September 2008
International
Telecommunication
42
Union
Lannion, France, 10-12 September 2008
Excitation
Masking curves
International
Telecommunication
43
Union
Lannion, France, 10-12 September 2008
Experiment 1
Loudness of synthesized noises
Physical parameters of the twelve synthesized noises
so u n d
n u m b e r
1
2
3
4
5
6
7
8
9
1 0
1 1
1 2
C e n tra l
fre q u e n c y
(H z )
4 0 0
4 0 0
4 0 0
4 0 0
1 4 2 0
1 4 2 0
1 4 2 0
1 4 2 0
3 0 0 0
3 0 0 0
3 0 0 0
3 0 0 0
b a n d w id th
(H z )
5 0
4 2 0
5 0
4 2 0
1 2 0
1 0 0 0
1 2 0
1 0 0 0
2 4 0
2 0 4 0
2 4 0
2 0 4 0
L e v e l
(d B S P L )
4
4
6
6
3
3
5
5
6
6
7
7
0
0
0
0
0
0
0
0
0
0
0
0
† 8 listeners
International
Telecommunication
44
Union
Lannion, France, 10-12 September 2008
From Meunier et al. InterNoise 2000
Experiment 1 : synthesized noises
(phons)
Adjusted loudness
y = 1,893 + 0,95909x R= 0,98632
90
80
70
60
50
40
30
11
Zwicker's model
49
3
12
10
6
5
2
7
30
8
1
40
50
60
Calculated loudness (phons)
70
80
90
(phons)
Adjusted loudness
y = 0,8725 + 0,9377x R= 0,99454
90
80
70
60
50
40
30
Moore's model (Acustica)
11 12
6
5
2
7
30
8
1
40
50
60
Calculated loudness (phons)
70
80
11
4
9
90
12
10
(phons)
y = 7,5766 + 0,85313x R= 0,97504
90
Moore's model (AES)
80
3
70
6
5
60
2
1
50
8
7
40
7
30
30
40
50
60
Calculated
loudness
(phons)
Lannion, France, 10-12 September 2008
Adjusted loudness
9 4 10
3
70
80
90
International
Telecommunication
45
Union
From Meunier et al. InterNoise 2000
Experiment 1
Loudness of environmental noises
Twenty four enrironmental sound (steady over 1 s)
Sound
† 24 listeners
Blowlamp
Guitare
Harmonica
Rumpled paper
Computer hard disk
Telephon in an Anecho•c Chamber
Telephon in an office
Bicycle in an Anecho•c Chamber
Bicycle
Car
Woman voice
Man voice
Flute at 39 dB SPL
Flute at 54 dB SPL
Flute at 69 dB SPL
Flute at 84 dB SPL
Motorcycle at 28 dB SPL
Motorcycle at 43 dB SPL
Motorcycle at 58 dB SPL
Motorcycle at 73 dB SPL
Drilling at 35 dB SPL
Drilling at 50 dB SPL
Drilling at 65 dB SPL
Drilling at 80 dB SPL
Lannion, France, 10-12 September 2008
Abbreviation
Blowlamp
Guitare
Harm
Paper
Disk
Tel_AC
Tel
Bicy_Ac
Bicy
Car
Voice_W
Voice_M
Flute_39
Flute_54
Flute_69
Flute_84
Moto_28
Moto_43
Moto_58
Moto
Drill_35
Drill_50
Drill_65
Drilling
International
Telecommunication
46
Union
From Meunier et al. InterNoise 2000
Experiment 2 : environmental noises
(phons)
Adjusted loudness
y = 15,611 + 0,77787x R= 0,98295
100
90
80
70
60
50
40
30
Zwicker's model
30
40
50
60
70
Calculated loudness (phons)
80
90
100
(phons)
Adjusted loudness
y = 3,6762 + 0,87937x R= 0,97893
100
90
80
70
60
50
40
Moore's model (Acustica)
40
50
60
70
Calculated loudness (phons)
80
90
100
(phons)
Adjusted loudness
y = 8,587 + 0,82236x R= 0,96235
100
90
80
70
60
50
40
Lannion, France, 10-12 September 2008
Moore's model (AES)
40
50
60
70
Calculated loudness (phons)
80
90
100
International
Telecommunication
47
Union
From Meunier et al. InterNoise 2000