RPI LRC Capturing the Lighting Edge New Color Metrics Mark Fairchild
Color Appearance of Displays, etc.
RGC scheduled September 28, 2012 from 6:45 AM to 9:45 AM RPI LRC Capturing the Lighting Edge New Color Metrics
Oct. 3, 2012
Mark Fairchild Rochester Institute of Technology, College of Science
1 Some Adaptation Demos
2 Color
3 4 5 6 Blur (Sharpness)
7 8 9 10 Noise
11 12 13 14 Chromatic, Blur, and Noise Adaptation
15 Outline
•Color Appearance Phenomena •Chromatic Adaptation •Metamerism •Color Appearance Models •HDR
16 Color Appearance Phenomena
17 Color Appearance Phenomena
If two stimuli do not match in color appearance when (XYZ)1 = (XYZ)2, then some aspect of the viewing conditions differs.
Various color-appearance phenomena describe relationships between changes in viewing conditions and changes in appearance.
Bezold-Brücke Hue Shift Abney Effect Helmholtz-Kohlrausch Effect Hunt Effect Simultaneous Contrast Crispening Helson-Judd Effect Stevens Effect Bartleson-Breneman Equations Chromatic Adaptation Color Constancy Memory Color Object Recognition
18 Simultaneous Contrast
The background in which a stimulus is presented influences the apparent color of the stimulus. Stimulus Indicates lateral interactions and adaptation.
Stimulus Color- Background Background Change Appearance Change
Darker Lighter
Lighter Darker
Red Green
Green Red
Yellow Blue
Blue Yellow
19 Simultaneous Contrast Example
(a)
(b)
20 Josef Albers
21 Complex Spatial Interactions
22 Hunt Effect
Corresponding chromaticities across indicated relative changes in luminance (Hypothetical Data) For a constant chromaticity, perceived 0.6 colorfulness increases with luminance.
0.5 As luminance increases, stimuli of lower colorimetric purity are required to match 1 10 a given reference stimulus. y 0.4 100 1000 10000
10000 1000 100 10 1 0.3 Indicates nonlinearities in visual processing.
0.2 0.2 0.3 0.4 0.5 0.6 x
23 Stevens Effect
1 Perceived lightness contrast increases with increasing adapting luminance.
As adapting luminance increases: dark colors look darker and light colors look lighter. L (Dark) L (63 dB) Indicates luminance-dependent nonlinearities. L (79 dB)
L (97 dB) 0.1 0.1 1 Y/Yn
24 Stevens & Hunt Effects
0.1 cd/m2 1.0 cd/m2 10 cd/m2
100 cd/m2 1000 cd/m2 10,000 cd/m2
25 Bartleson-Breneman
Perceived lightness as a function of relative luminance for various surround relative luminances 100 90 Apparent contrast in complex stimuli (i.e. images) 80 increases with increasing surround luminance. 70 60 Decreased surround luminance increases the 50 brightness of all image colors, but the effect is 40 greater for dark colors. 30 Lightness (Average) 20 Lightness (Dim) Indicates a differential contrast effect (white-point 10 Lightness (Dark) resetting). 0 0 10 20 30 40 50 60 70 80 90 100 Relative Luminance, Y
26 Surround Effect Demo
27 28 Adaptation
Light Adaptation: Decrease in visual sensitivity with increases in luminance. (Automatic Exposure Control)
Dark Adaptation: Increase in visual sensitivity with decreases in luminance. (Automatic Exposure Control)
Chromatic Adaptation: Independent sensitivity regulation of the mechanisms of color vision. (Automatic Color Balance)
29 Local Adaptation
Linear Mapping Perceptual Mapping
30 Chromatic Adaptation
1.25 The three cone types, LMS, are capable of 1 independent sensitivity regulation. (Adaptation occurs in higher-level mechanisms as well.) 0.75 Magnitudes of chromatic responses are dependent 0.5 on the state of adaptation (local, spatial, temporal). Afterimages provide evidence. 0.25
0 400 450 500 550 600 650 700 Wavelength (nm)
31 32 A Cyan Filter
33 34 35 Color Constancy (Discounting)
0.54 Adapting Chromaticities Achromatic Chromaticities Incandescent We perceive the colors of objects to remain 0.52 unchanged across large changes in illumination <-- Hands <-- No Hands color. •Not True v' 0.50 •Chromatic Adaptation •Poor Color Memory •Cognitive Discounting-the-Illuminant 0.48 Daylight
<-- Hands <-- No Hands 0.46 0.18 0.20 0.22 0.24 0.26
u'
36 Diamonds
37 With Tips
38 On Backgrounds
39 Both Tips & Backgrounds!
40 Real-World Discounting
41 42 Chromatic Adaptation Modeling
Chromatic Adaptation: Largely independent sensitivity regulation of the (three) mechanisms of color vision.
1.25
1
0.75
0.5
0.25
0 400 450 500 550 600 650 700 Wavelength (nm) 43 Chromatic Adaptation Models Model: X1Y1Z1 La = f(L, Lwhite ,...) 3x3 Ma = f(M, Mwhite,...) L1M1S1 S a = f(S, Swhite,...) VC1
LaMaSa
Transform (CAT): VC2 L M S XYZ 2 = f(XYZ1, XYZwhite ,...) 2 2 2 3x3
X2Y2Z2
44 XYZ-to-LMS
2 z 1
S M L 1.5
y 0.75 s t i e v i u l t i a s L 0.400 0.708 −0.081 X v
n x
e y s s u
0.5 1 l e u v i m t i t a l M = −0.226 1.165 0.046 Y s i e r R 0.25 T S 0.000 0.000 0.918 Z 0.5 0 380 480 580 680 780 Wavelength, nm 0 380 480 580 680 780 Wavelength, nm
45 Johannes von Kries
Johannes von Kries "Father of Chromatic-Adaptation Models"
46 von Kries Hypothesis
"This can be conceived in the sense that the individual components present in the organ of vision are completely independent of one another and each is fatigued or adapted exclusively according to its own function." -von Kries, 1902
La = kLL
Ma = k MM
S a = kSS
Von Kries thought of this “proportionality law” as an extension of Grassmann’s Laws to span two viewing conditions.
47 Some Evolution of CATs •Nayatani et al. Nonlinear Model •Fairchild (1991) & (1994) Models •Bradford Model •CIELAB & CIELUV •CAT02
48 Metamerism: Illuminant, Observer, & Metrics
49 Illuminant Metamerism
• Same Observer, Change Illuminantion • In theory, must have tristimulus equality for a reference illuminant. • Calculate Color Difference for Test Illuminant: CIE Special Index of Metamerism • Calculate Spectral Difference: General Index of Metamerism
50 In Practice, Paramers, not Metamers
Corrected Standard Batch batch
X 19.93 21.96 19.93
Y 14.29 14.87 14.29
Z 7.23 7.91 7.23
0.8 HJ HJ HJ J 0.6 H r o t Standard
c B a f
e J J Batch c
n 0.4 H a t B H Corrected batch c
e B l f B e B
R B J B 0.2 H B HJ HJ H H J B B BJ H J J BHJ H H BHJ BHJ BJ B B 0 400 450 500 550 600 650 700 Wavelength (nm) 51 CIE Special Index of Metamerism, Change in Illuminant
• Must have an exact tristimulus match for a single viewing and observer condition. • Calculate Color Difference for Test Illuminant • Workflow: • Measure Parameric Pair • Correct Batch as Needed for Reference Illuminant • Calculate CIE DE2000 for Test Illuminant
52 General Index of Metamerism is Spectral Based
• Weighted RMS Error (Using Diagonal of Matrix R)
n ⎛ ⎞2 ⎜ w ΔR ⎟ ∑⎝ λ λ ⎠ WRMS = λ=1 n w = diag R λ ( )
53 Observer Metamerism 1931 2° vs. 1964 10° standard observers
2
1.5 s e u l a v
s
u 1 l u m i t s i r T 0.5
0 380 480 580 680 780 Wavelength, nm
Dashed lines, 1964 observer
54 Variability in Color Matching Functions
Wyszecki and Stiles, “Color Science” 55 CIE Standard Deviate Observer
2.2 • Based on Stiles and Burch 1.8 data s e
u 1.4 l a v
s Normalization of primaries
u 1 l • u and V-lambda minimized m i t s
i 0.6 r true variance T 0.2 Not recommended -0.2 • 380 480 580 680 780 Wavelength, nm • CIE Pub. 80 20-year old (solid lines) 60-year old (dashed lines)
56 CIE TC 1-36 (2006)
⎣⎡− Dτ ,max,macula ⋅Dmacula,relative (λ)− Dτ ,ocul (λ)⎦⎤ l (λ) = αi,l (λ)⋅10
⎣⎡− Dτ ,max,macula ⋅Dmacula,relative (λ)− Dτ ,ocul (λ)⎦⎤ m(λ) = αi,m (λ)⋅10
⎣⎡− Dτ ,max,macula ⋅Dmacula,relative (λ)− Dτ ,ocul (λ)⎦⎤ s (λ) = αi,s (λ)⋅10
Color-matching Cone spectral Macula spectral Lens and ocular functions absorptances density media spectral as a function of density as a field size function of age (i=internal)
57 Examples of Cone Fundamentals
L-, M-, & S-Cone Fundamentals (2- & 10-Deg. @ Age 32) L-, M-, & S-Cone Fundamentals (10-Deg. @ Ages 20 & 80)
1.00 1.00 l-bar (2) l-bar (20) 0.90 m-bar (2) 0.90 m-bar (20) s-bar (2) s-bar (20) 0.80 l-bar (10) 0.80 l-bar (80) m-bar (10) m-bar (80) 0.70 0.70 s-bar (10) s-bar (80)
0.60 0.60
0.50 0.50
0.40 0.40
Relative Sensitivity 0.30 Relative Sensitivity 0.30
0.20 0.20
0.10 0.10
0.00 0.00 390 440 490 540 590 640 690 740 390 440 490 540 590 640 690 740 Wavelength (nm) Wavelength (nm)
Normalized to Peak Height
58 Effect of Age vs. Field Size
2 2 xbar 1931 xbar 1931 ybar 1931 ybar 1931 zbar 1931 zbar 1931 x25yrs,2degree x25,10degree 1.5 1.5 y25yrs,2degree y25,10degree z25yrs,2degree z25,10degree x65yrs,2degree x65,10degree y65yrs,2degree y65,10degree 1 1 z65yrs,2degree z65,10degree ristimulus values ristimulus values
T 0.5 T 0.5
0 0
-0.5 -0.5 380 430 480 530 580 630 680 730 780 380 430 480 530 580 630 680 730 780 Wavelength (nm) Wavelength (nm)
59 Some Metamers
0.7 0.8 Vinyl-1 O Vinyl-1 G 0.7 0.6 Vinyl 2-A Vinyl 2-B 0.6 Fabric-A 0.5 Fabric-B
0.5 0.4
0.4
0.3 Reflectance factor Reflectance factor 0.3
0.2 0.2
0.1 0.1
0 0 380 430 480 530 580 630 680 730 780 380 430 480 530 580 630 680 730 780 Wavelength (nm) Wavelength (nm)
Grays Materials
60 Observer Metamerism
2,25 2,65
1931
10,25 61 10,65 Illuminant Metamerism
1931, D65 1931, A
1931, HP Na 1964, HP Na
62 Color “Blindness”
63 Cone Spectral Sensitivities
1
S M L
y 0.75 t i v i t i s n e s
0.5 e v i t a l e
R 0.25
0 380 480 580 680 780 Wavelength, nm
64 Color Deficiencies
Dichromats
Protanopia: Lacking “L-cone response” Deuteranopia: Lacking “M-cone response” Tritanopia: Lacking “S-cone response” Also “Anomalous Trichromats” ~8% of Males have Color Vision Deficiencies
Various tests available at:
“Normal” Deuteranope
66 Color Vision Deficiencies
Original
Protanope (Missing L Cones)
Pseudo-isochromatic plates Dueteranope (Missing M Cones)
Tritanope (Missing S Farnsworth- Cones) Munsell 100-Hue vischeck.com Test 67 Chromatic Adaptation Reminder
68 Definitions • Chromatic adaptation: changes in the visual system that approximately compensate for changes in the spectral quality of illumination. • Light adaptation: changes in the visual system that approximately compensate for changes in the level of illumination.
Luminance
0.0001 .001 .1 to 1 5 20 50 to 150 150 to 600 3000 6000 +10000 Luminance (cd/m2)
69 Definitions
• Corresponding color stimuli: Pairs of color stimuli that look alike when one is seen in one set of adaptation conditions and the other is seen in a different set. • Chromatic adaptation transform (CAT) – computational approach to calculate corresponding colors,
70 Corresponding Color Experiment
A color chip is selected under daylight that matches the perception of the sample viewed under tungsten
71 Color Inconstancy Index
72 Three-Color Gray is Color Inconstant
0.5 0.5
0.4 0.4
0.3 0.3 Factor Factor
0.2 0.2 Reflectance 0.1 Reflectance 0.1
0.0 0.0 380 440 500 560 620 680 380 440 500 560 620 680 Wavelength (nm) Wavelength (nm)
Daylight 73 Incandescent Color Inconstancy (Left) vs. Metamerism (Right)
One Object Two Objects
74 Corresponding Color Calculation
⎛ X ⎞ ⎛ X⎞ ⎜ c⎟ ⎜ ⎟ −1 ⎜ Yc ⎟ = MCAT02MVKMCAT02⎜ Y⎟ ⎜ Z ⎟ ⎜ Z⎟ ⎝ c ⎠ ⎝ ⎠