Col
Content
The Color Eye.
Part 1: Human color vision
Human color vision, KIG/Macbeth, G. Rösler 643
Page 1
Object colors
4. "Color" 1. Illumination 3. Observer
2. Sample
Human color vision, KIG/Macbeth, G. Rösler 643
Page 2
Illumination
Sunlight: The fundamental reason for vision on earth
Electromagnetic spectrum: Visible range is very small
Gamma radiation Ultraviolet radiation visible radiation Infrared radiation Radar, TV Radio Audio Frequencies Energy distribution X-Rays
Kosmic radiation
1pm
1nm
1um
1mm
1m
1km
1Mm
1Gm
Wavelength
380nm
730nm
Human color vision, KIG/Macbeth, G. Rösler 643
Page 3
Natural and artifical lights: D65, A, TL 84
CIE Standard Illuminants
D65 (Daylight) A (Tungsten) F11 (Fluorescent)
Rel. Intensity
300
400
500
600
700
800
Wavelength (nm)
Human color vision, KIG/Macbeth, G. Rösler 643 Page 4
Objects
Reflection: What we see with our eyes Absorption: That is what colorants do. Transmission, opacity, hiding power: Sometimes important
Human color vision, KIG/Macbeth, G. Rösler 643
Page 5
Eye
Functional description
Our "camera" design Spatial resolution Color resolution
Eye Ball Lens
Retina with 120.000.000 receptors
Light
800.000 nerves to the brain
Human color vision, KIG/Macbeth, G. Rösler 643
Page 6
Sensitivity
Two types of sensors Low light levels: Rods (120 Million) High density all over the retina, but no reds in the center, grey vision Normal light levels: 3 types of Cones (6 Million) Low density all over the retina, but very high density in the center, color vision Color vision deficiencies : Missing cone signals
Page 7
Human color vision, KIG/Macbeth, G. Rösler 643
Human color vision, KIG/Macbeth, G. Rösler 643
Page 8
Human color vision, KIG/Macbeth, G. Rösler 643
Page 9
Human color vision, KIG/Macbeth, G. Rösler 643
Page 10
Human color vision, KIG/Macbeth, G. Rösler 643
Page 11
Cone spectral responses
Standard observer x,y,z, Field of view 2°, 10°
Standard observer 2° field of view
2,5
2
Standard observer 10° field of view
2,5
rel. sensitivity
1,5
1
2
rel. sensitivity
300 400 500 600 700
0,5
1,5
0
1
Wavelength (nm)
0,5
0 300 400 500 600 700
Wavelength (nm)
Human color vision, KIG/Macbeth, G. Rösler 643
Page 12
Brain
The color sensation is generated in the brain Complex vision processing is happening More research necessary in order to understand the information processing and to develop better mathemetical models for instrumental color metrics.
Human color vision, KIG/Macbeth, G. Rösler 643
Page 13
How are colors seen
Context
Surround, contrast
(Center colour is the same, only surround is different)
Human color vision, KIG/Macbeth, G. Rösler 643
Page 14
Perceived colors are influenced by
Size Adaptation Natural scenes Artificial scenes Structure Expectation Experience Discipline
Human color vision, KIG/Macbeth, G. Rösler 643 Page 15
Color names
Verbal names Color collections
e.g. RAL Munsell
Artist, uniform color space Basis of CIE Lab
Modern color order systems
Human color vision, KIG/Macbeth, G. Rösler 643
Page 16
Visual comparison of colors
What is possible?
What we see is what counts How many colors can be differentiated? Several Million colors !!!!!
What is necessary?
Standardised illumination Standardised viewing geometries
Human color vision, KIG/Macbeth, G. Rösler 643
Page 17
Visual comparison of colors
What is not possible?
Limitations
Side by side necessary Memory not accurate enough
Human color vision, KIG/Macbeth, G. Rösler 643
Page 18
Surprises
Metamerism
Daylight Tungsten or some other illumination Sample 1
Sample 1
similar
Sample 2 Metamerism
different
Sample 2
Human color vision, KIG/Macbeth, G. Rösler 643
Page 19
Human color vision, KIG/Macbeth, G. Rösler 643
Page 20
Human color vision, KIG/Macbeth, G. Rösler 643
Page 21
Human color vision, KIG/Macbeth, G. Rösler 643
Page 22
Human color vision, KIG/Macbeth, G. Rösler 643
Page 23
Human color vision, KIG/Macbeth, G. Rösler 643
Page 24
Surprises
Directionality
A B A
B
Structures
B A B A
Human color vision, KIG/Macbeth, G. Rösler 643
Page 25
Surprises
Effect colors
Metallic effect Interference effect
Multiple viewing geometries are necessary Directed viewing and directed illumination give best result
Page 26
Human color vision, KIG/Macbeth, G. Rösler 643
Surprises
Fluorescence
Optical brightner Popular in paper and textiles Adds "blue" to the yellowish substrate for "white" appearance. Only works with illumination containing UV light Daylight fluorescence e.g. red, green. Looks lighter and has higher chroma than other objects with the same illumination.
Human color vision, KIG/Macbeth, G. Rösler 643
Page 27
Color identification
The "address" is valid for a certain illuminant (D65 or A or TL84 or others) and for a certain observer (2° or 10°) and for a certain geometry (d/8°, 45°/0° or others) ! Changing illuminant and/or observer and/or geometry in most cases will result in a different "address" in color space for the same sample
Human color vision, KIG/Macbeth, G. Rösler 643
Page 28
Color identification
Color order system, color space
CIE L*a*b* 1976 Based on the Munsell system L* a* b* are the primary coordinates L*: Lightness direction a*: red - green direction b*: yellow - blue direction
CIE Lab system is industry standard
Human color vision, KIG/Macbeth, G. Rösler 643
Page 29
L* Lightness
L*= 100 white
L*= 50 grey
L*= 0 black
Human color vision, KIG/Macbeth, G. Rösler 643
Page 30
a* red - green b* yellow - blue
b*= + xx yellow
a*= - xx green
a*= + xx red
b*= + xx blue
Human color vision, KIG/Macbeth, G. Rösler 643 Page 31
CIE Lab system version 2
C*, h: Same facts, different description polar coordinates instead of kartesian coordinates Chroma (independent of hue) C*= (a*2 + b*2) 1/2 Hue angle (independent of chroma) h= arctan (b* / a*)
counterclockwise, starting at positive a* axis with h = 0°
Human color vision, KIG/Macbeth, G. Rösler 643
Page 32
h=90°
C* Chroma h hue angle
h=180°
h=0°
C* = + x low chroma
h=270°
Human color vision, KIG/Macbeth, G. Rösler 643
C* = + xx high chroma
Page 33
Color difference description
Most important for industrial use Color difference is the distance in color space between two "addresses", e.g. between the "address" of a standard and the "address" of a trial. The color difference has 3 components: Lightness difference dL*=L*trial - L*standard red-green difference da*=a*trial - a*standard yellow-blue difference db*=b*trial - b*standard
Page 34
Human color vision, KIG/Macbeth, G. Rösler 643
Total color difference dE* is calculated from the 3 components: dE*= (dL*2 + da*2 + db*2) 1/2 Same facts, different description dE*= (dL*2 + dC*2 + dH*2) 1/2 Chroma difference dC* = C*trial - C*standard Hue difference dH*= (da*2 + db*2 - dC*2) 1/2
sign dH*: positive if a*trial b*standard - a*standard b*trial R0 else negative.
Attention: dH* is a distance in color space and not the differenec of the hue angles htrial - hstandard !!! But the computer has no problem with the slightly complicated calculation
Human color vision, KIG/Macbeth, G. Rösler 643
Page 35
Communications
Physical standards
Generating standards Stability of standards Regular comparison of standards
Human color vision, KIG/Macbeth, G. Rösler 643
Page 36
Part 1: Human color vision
Human color vision, KIG/Macbeth, G. Rösler 643
Page 1
Object colors
4. "Color" 1. Illumination 3. Observer
2. Sample
Human color vision, KIG/Macbeth, G. Rösler 643
Page 2
Illumination
Sunlight: The fundamental reason for vision on earth
Electromagnetic spectrum: Visible range is very small
Gamma radiation Ultraviolet radiation visible radiation Infrared radiation Radar, TV Radio Audio Frequencies Energy distribution X-Rays
Kosmic radiation
1pm
1nm
1um
1mm
1m
1km
1Mm
1Gm
Wavelength
380nm
730nm
Human color vision, KIG/Macbeth, G. Rösler 643
Page 3
Natural and artifical lights: D65, A, TL 84
CIE Standard Illuminants
D65 (Daylight) A (Tungsten) F11 (Fluorescent)
Rel. Intensity
300
400
500
600
700
800
Wavelength (nm)
Human color vision, KIG/Macbeth, G. Rösler 643 Page 4
Objects
Reflection: What we see with our eyes Absorption: That is what colorants do. Transmission, opacity, hiding power: Sometimes important
Human color vision, KIG/Macbeth, G. Rösler 643
Page 5
Eye
Functional description
Our "camera" design Spatial resolution Color resolution
Eye Ball Lens
Retina with 120.000.000 receptors
Light
800.000 nerves to the brain
Human color vision, KIG/Macbeth, G. Rösler 643
Page 6
Sensitivity
Two types of sensors Low light levels: Rods (120 Million) High density all over the retina, but no reds in the center, grey vision Normal light levels: 3 types of Cones (6 Million) Low density all over the retina, but very high density in the center, color vision Color vision deficiencies : Missing cone signals
Page 7
Human color vision, KIG/Macbeth, G. Rösler 643
Human color vision, KIG/Macbeth, G. Rösler 643
Page 8
Human color vision, KIG/Macbeth, G. Rösler 643
Page 9
Human color vision, KIG/Macbeth, G. Rösler 643
Page 10
Human color vision, KIG/Macbeth, G. Rösler 643
Page 11
Cone spectral responses
Standard observer x,y,z, Field of view 2°, 10°
Standard observer 2° field of view
2,5
2
Standard observer 10° field of view
2,5
rel. sensitivity
1,5
1
2
rel. sensitivity
300 400 500 600 700
0,5
1,5
0
1
Wavelength (nm)
0,5
0 300 400 500 600 700
Wavelength (nm)
Human color vision, KIG/Macbeth, G. Rösler 643
Page 12
Brain
The color sensation is generated in the brain Complex vision processing is happening More research necessary in order to understand the information processing and to develop better mathemetical models for instrumental color metrics.
Human color vision, KIG/Macbeth, G. Rösler 643
Page 13
How are colors seen
Context
Surround, contrast
(Center colour is the same, only surround is different)
Human color vision, KIG/Macbeth, G. Rösler 643
Page 14
Perceived colors are influenced by
Size Adaptation Natural scenes Artificial scenes Structure Expectation Experience Discipline
Human color vision, KIG/Macbeth, G. Rösler 643 Page 15
Color names
Verbal names Color collections
e.g. RAL Munsell
Artist, uniform color space Basis of CIE Lab
Modern color order systems
Human color vision, KIG/Macbeth, G. Rösler 643
Page 16
Visual comparison of colors
What is possible?
What we see is what counts How many colors can be differentiated? Several Million colors !!!!!
What is necessary?
Standardised illumination Standardised viewing geometries
Human color vision, KIG/Macbeth, G. Rösler 643
Page 17
Visual comparison of colors
What is not possible?
Limitations
Side by side necessary Memory not accurate enough
Human color vision, KIG/Macbeth, G. Rösler 643
Page 18
Surprises
Metamerism
Daylight Tungsten or some other illumination Sample 1
Sample 1
similar
Sample 2 Metamerism
different
Sample 2
Human color vision, KIG/Macbeth, G. Rösler 643
Page 19
Human color vision, KIG/Macbeth, G. Rösler 643
Page 20
Human color vision, KIG/Macbeth, G. Rösler 643
Page 21
Human color vision, KIG/Macbeth, G. Rösler 643
Page 22
Human color vision, KIG/Macbeth, G. Rösler 643
Page 23
Human color vision, KIG/Macbeth, G. Rösler 643
Page 24
Surprises
Directionality
A B A
B
Structures
B A B A
Human color vision, KIG/Macbeth, G. Rösler 643
Page 25
Surprises
Effect colors
Metallic effect Interference effect
Multiple viewing geometries are necessary Directed viewing and directed illumination give best result
Page 26
Human color vision, KIG/Macbeth, G. Rösler 643
Surprises
Fluorescence
Optical brightner Popular in paper and textiles Adds "blue" to the yellowish substrate for "white" appearance. Only works with illumination containing UV light Daylight fluorescence e.g. red, green. Looks lighter and has higher chroma than other objects with the same illumination.
Human color vision, KIG/Macbeth, G. Rösler 643
Page 27
Color identification
The "address" is valid for a certain illuminant (D65 or A or TL84 or others) and for a certain observer (2° or 10°) and for a certain geometry (d/8°, 45°/0° or others) ! Changing illuminant and/or observer and/or geometry in most cases will result in a different "address" in color space for the same sample
Human color vision, KIG/Macbeth, G. Rösler 643
Page 28
Color identification
Color order system, color space
CIE L*a*b* 1976 Based on the Munsell system L* a* b* are the primary coordinates L*: Lightness direction a*: red - green direction b*: yellow - blue direction
CIE Lab system is industry standard
Human color vision, KIG/Macbeth, G. Rösler 643
Page 29
L* Lightness
L*= 100 white
L*= 50 grey
L*= 0 black
Human color vision, KIG/Macbeth, G. Rösler 643
Page 30
a* red - green b* yellow - blue
b*= + xx yellow
a*= - xx green
a*= + xx red
b*= + xx blue
Human color vision, KIG/Macbeth, G. Rösler 643 Page 31
CIE Lab system version 2
C*, h: Same facts, different description polar coordinates instead of kartesian coordinates Chroma (independent of hue) C*= (a*2 + b*2) 1/2 Hue angle (independent of chroma) h= arctan (b* / a*)
counterclockwise, starting at positive a* axis with h = 0°
Human color vision, KIG/Macbeth, G. Rösler 643
Page 32
h=90°
C* Chroma h hue angle
h=180°
h=0°
C* = + x low chroma
h=270°
Human color vision, KIG/Macbeth, G. Rösler 643
C* = + xx high chroma
Page 33
Color difference description
Most important for industrial use Color difference is the distance in color space between two "addresses", e.g. between the "address" of a standard and the "address" of a trial. The color difference has 3 components: Lightness difference dL*=L*trial - L*standard red-green difference da*=a*trial - a*standard yellow-blue difference db*=b*trial - b*standard
Page 34
Human color vision, KIG/Macbeth, G. Rösler 643
Total color difference dE* is calculated from the 3 components: dE*= (dL*2 + da*2 + db*2) 1/2 Same facts, different description dE*= (dL*2 + dC*2 + dH*2) 1/2 Chroma difference dC* = C*trial - C*standard Hue difference dH*= (da*2 + db*2 - dC*2) 1/2
sign dH*: positive if a*trial b*standard - a*standard b*trial R0 else negative.
Attention: dH* is a distance in color space and not the differenec of the hue angles htrial - hstandard !!! But the computer has no problem with the slightly complicated calculation
Human color vision, KIG/Macbeth, G. Rösler 643
Page 35
Communications
Physical standards
Generating standards Stability of standards Regular comparison of standards
Human color vision, KIG/Macbeth, G. Rösler 643
Page 36
