![Brian Kim
Light (electromagnetic radiation)
5
Wavelength (λ) and frequency (v)
o λ=c / v ( c is the speed of light)
Electromagnetic Energy and Frequency
o E=hv=hλ/c (h is Planck’s constant)
[Unit: λ: meters; v: Hertz; E: electron-volt]
Visible spectrum: ~ 400 to 700 nm
o Electromagnetic radiation
o “Light” usually refers to radiation visible to
human eye
Properties of EM](https://image.slidesharecdn.com/introductiontocolorscience10-230930174515-165a17a2/85/Introduction-to-Color-Science-for-display-engineer-5-320.jpg)
![Brian Kim
Luminance vs. Brightness
Luminance
o Measured amount of light in [cd/m2]
o Physical quantity
-the light energy weighted by the spectral sensitivity function
of the human visual system
-independent of the luminance of the surrounding objects
Brightness
o Perceived amount of light (perceived luminance)
o Psychophysical quantity
-depends on luminance of the surround (lateral inhibition,
contrast)
34
Brightness
Luminance ](https://image.slidesharecdn.com/introductiontocolorscience10-230930174515-165a17a2/85/Introduction-to-Color-Science-for-display-engineer-34-320.jpg)
![Brian Kim
Relationship between Radiance and Luminance
35
Luminance – how bright the surface will appear regardless of its color. [units: cd/m2]
o Photometric quantity defined by the spectral luminous efficiency function
o L ≈ 0.2126 R + 0.7152 G + 0.0722 B
Luminance Light spectrum
(radiance)
Luminous efficiency function
(weighting)
= x
Luminosity curve = Eye response
curve = Luminous efficiency curve](https://image.slidesharecdn.com/introductiontocolorscience10-230930174515-165a17a2/85/Introduction-to-Color-Science-for-display-engineer-35-320.jpg)
Introduction to Color Science for display engineer
- 1. Introduction to Color Science AUG.7, 2023 BRIAN KIM 1 Please note that while this article is based on public information, I cannot guarantee that any illustrations cited may not be in violation of copyright.
- 2. Brian Kim Contents 2 1. How does the human eye detect color? 2. How was the CIE color space created? CIE (International Commission on Illumination) 3. Underlying physics of CIE color coordinates. 4. How to choose a light source for a display?
- 3. Brian Kim 1. How does the human eye detect color? 3
- 4. Brian Kim Object color & Source color 4 Object color: reflect or transmit the light from light from the source. Source color: directly emitted from light source. Pointer’s DCI-P3 Object color Source color
- 5. Brian Kim Light (electromagnetic radiation) 5 Wavelength (λ) and frequency (v) o λ=c / v ( c is the speed of light) Electromagnetic Energy and Frequency o E=hv=hλ/c (h is Planck’s constant) [Unit: λ: meters; v: Hertz; E: electron-volt] Visible spectrum: ~ 400 to 700 nm o Electromagnetic radiation o “Light” usually refers to radiation visible to human eye Properties of EM
- 6. Brian Kim The human visual system 6 Human eye does not directly measure the spectrum of incoming light a.k.a. the brain does not receive “a spectrum” from the eye The eye measures three response values = (S, M, L). The result of integrating the incoming spectrum against response functions of S, M, L-cones
- 7. Brian Kim Human Vision 7 Retina consists of an array of rods and three kinds of cones o Rods are for night vision o Cones are for color vision Three kinds of cones: 1. L-cone: most sensitive to red light 2. M-cone: most sensitive to green light 3. S-cone: most sensitive to blue light Source: http://webvision.med.utah.edu/sretina.html
- 8. Brian Kim Distribution of Rods and Cones 8 Scotopic vision mainly at outer parts of the retina Photopic and color vision centered at the fovea (by far highest visual acuity) No receptors at blind spot (optic nerve) (코 쪽을 nasal retina, 측두엽 쪽을 temporal retina)
- 9. Brian Kim Scotopic and photopic: eye response 9 Dark-adapted (scotopic) vision is sensitive to shorter wavelengths, while light-adapted (photopic) vision is sensitive to all visible wavelengths, although not equally. The curves represent the spectral luminous efficacy for human vision. o The lumen is defined such that the peak of the photopic vision curve has a luminous efficacy of 683 lumens/watt. o The response curve of the eye along with the spectral power distribution of a luminous object determine the perceived color of the object.
- 10. Brian Kim The eye’s photoreceptor cells: rods & cones 10 Cones (photopic vision) Rods (scotopic vision) • 5~8 millions • Central vision • High density in the macula and fovea • Less sensitive to light • Three types • Color vison • 100~200 Millions • Peripheral vision • Located everywhere except fovea • Very sensitive to light situations • One type • Black and white
- 11. Brian Kim More about cone (trichromatic vision) 11 S-cones (S for short wavelength) o Most sensitive to blue light (~430 nm) o Only 5-6% of the cones are S-cones M-cones (M for medium wavelength) o Most sensitive to green light (~ 530 nm) L-cones (L for long wavelength) o Most sensitive to yellow-green light (~560 nm) o Ratio of L/M cones highly varies for different individuals
- 12. Brian Kim Eye sensitivity function, V() 12 The human eye has maximum conversion efficiency at 555 nm, meaning that green has maximum visual sensitivity. By utilizing this characteristic, the diamond pentile pixel structure places two greens in one pixel. This allows low-resolution displays to have high-resolution visual characteristics
- 13. Brian Kim 2. How was the CIE color space created? 13
- 14. Brian Kim Color-Matching Experiments 14 Experimental Determination of Color-Matching Functions o Adjust amounts of primary lights to match a monochromatic test light o Not all test lights could be matched -Possibly to move any of the primaries to the side of the test light (counted as negative value)
- 15. Brian Kim 15 Amounts of the red, green and blue primaries needed to match any color The amounts of R, G, and B the subject selects to match each single-wavelength light forms the color-matching curves. o These are denoted ̅( ), ̅( ), ̅ ( ). What's a negative amount of color??? The negative part of the curve Indicates that some color cannot be reproduced by a linear combination of the primaries. For such color, one or more primary lights has to be shifted from one side to the other CIE 1931 RGB Color-Matching Functions Obtained using experimental data of Wright and Guild Basis of today’s color specification (CIE 1931 Standard Colorimetric Observer)
- 16. Brian Kim CIE Color Matching Functions 16 The CIE has proposed a new color matching function that is mathematically modified to avoid negative values. CIE XYZ Matching Functions CIE RGB Matching Functions RGB to XYZ
- 17. Brian Kim CIE 1931 XYZ Color-Matching Functions 17 Formed nonlinearly from X, Y and Z x and y are chromaticity coordinates, Y is relative luminance coordinate Gamut o Set of colors that can be physically realized in a color space (without negative coefficients etc.) Amounts of the XYZ primaries needed to match any color (y function is precisely CIE-standardized photopic luminous efficiency, 1931) 𝑋 = 𝐸 (𝜆) 𝑥̄(𝜆) 𝑑𝜆 𝑌 = 𝐸 (𝜆) 𝑦̄(𝜆) 𝑑𝜆 𝑍 = 𝐸 (𝜆) 𝑧̄(𝜆) 𝑑𝜆
- 18. Brian Kim CIE space 18 How to derive (x,y) values for a light source y-axis x-axis https://www.radiantvisionsystems.co m/products
- 19. Brian Kim Instrument for color management 19 x Color filter Grating Physics Measurement Output
- 20. Brian Kim 3. Underlying physics of CIE color coordinates. 20
- 21. Brian Kim Color Space Color gamut o The horse shoe area - Represents all the colors that human eye can see o The Pointers gamut is - The real surface colors can be seen by the human eye, based on the research by Michael R. Pointer (1980). o D65, Standard White Color - The color of outdoor at noon with color temperature close to 6500K CIE 1931 vs. CIE 1976 o CIE 1931 - Very non-uniform and distorted representation of human eye perception - Equal distance ≠ equal perceived color distance o CIE 1976 - Perceptually uniform for human color vision 21 DCI-P3 compared with Pointers gamut https://tftcentral.co.uk/articles/pointers_gamut https://www.yujiintl.com/understanding-cie1931-and-cie-1976/
- 22. Brian Kim Display Characteristic: Chromaticity Gamut Area 22 Color gamut: Color gamut represents the color range that can be reproduced artificially. o CIE 1931 xy : Most basic color space defined by CIE → Perceptually nonuniform o CIE 1976 u’v’ : Improve color scale to be more perceptually uniform. Both chromaticity values can be calculated from XYZ tristimulus values x = X / (X+Y+Z), y = Y / (X+Y+Z) u’ = 4X / (X+15Y+3Z), v’ = 9Y / (X+15Y+3Z)
- 23. Brian Kim CIE Chromaticity Diagram 23 Non spectral purples Spectral Locus Monochromatic wavelengths on outside curve (horseshoe) Planckian Locus (Blackbody curve) Defines color temperature of a source (Correlated Color Temperature, CCT, given in Kelvin) Color Gamut Range of colors that can be rendered by a device using the device’s primaries. R+G+B Color Space Every visible color to the human eye. Colors given in coordinates (x,y) Cannot displayed on normal display.
- 24. Brian Kim CIE Chromaticity Diagram 24 Dominant wavelengths or hues go around the perimeter of the chromaticity diagram. o A color’s dominant wavelength is where a line from white through that color intersects the perimeter. o Some colors, called non-spectral color’s, don’t have a dominant wavelength. Excitation purity or saturation is measured in terms of a color’s position on the line to its dominant wavelength. Complementary colors lie on opposite sides of white, and can be mixed to get white. A B Dominant wavelength Purity = B/(A+B) White Complementary color of Blue primary
- 25. Brian Kim 25 Peak Wavelength o Peak wavelength is defined as the single wavelength where the radiometric emission spectrum of the light source reaches its maximum. - More simply, it does not represent any perceived emission of the light source by the human eye, but rather by photo-detectors. Dominant Wavelength o Dominant wavelength is defined as the single wavelength that is perceived by the human eye. - Generally one light source consists of multiple wavelength spectrums from the light source rather than one single wavelength. Our brains turn those multiple spectrums into a single color of light consistent with a single specific wavelength which is what we see when we look at the light. That’s the light source’s Dominant wavelength. In general, these two parameters are not drastically different, but it can pay to consider our application when using these two parameters. o For example, if the LED is used in optical instruments and machines are being used to identify the wavelength, you should use Peak Wavelength for your LED selection. o If the LED is used to backlight a display or otherwise illuminate or indicate something for human operators, you should use Dominant Wavelength for your LED selection. Peak Wavelength vs. Dominant Wavelength
- 26. Brian Kim 4. How to choose a light source for a display? 26
- 27. Brian Kim An evolution of display system and color gamut NTSC NTSC REC 709 sRGB REC 709 sRGB Adobe RGB Adobe RGB DCI-P3 DCI-P3 The first official Color Gamut Standard for the beginning of US color television broadcasting. Too much saturated, not realistic 2014 Sochi Winter Olympic For printing industry by Adobe system, 96% NTSC 17% larger than sRGB/Rec709 For digital cinema production by the Digital Cinema Initiatives (DCI) organization . 26% larger than sRGB/Rec709 ITU Recommendation, the standards for UHDTV (UHD 4K and UHD 8K) based on NHK’s Super Hi-Vision •Intended for RGB lasers •Covers ~100% of Pointer’s Gamut •99.9% of Pointer’s gamut 72% larger than sRGB/Rec709 2020 Tokyo Olympic sRGB:PC standard by IEC Rec.709: HDTV standard Same 3 primary color with different curve 71% of NTSC 27 http://www.nhk.or.jp/8k/index_e.html# 1941 NTSC standards are set for B&W TV 1939 RCA presents TV at New York World’s Fair 1948 Television in homes gradually transitioned from black- and-white to color television between 1953 and 1974. 1964 NHK begins HDTV development 1999 Stations begin broadcasting Digital TV and wide screen HDTV. 2009 Broadcast TV in U.S. goes all digital 1988 test HD TV broadcasting in Japan 2016 Rio Olympic 2018 Pyeongchang Winter Olympic 2004 3 million HDTV sets in homes. 1950 Korean War slows color TV progress REC 2020 REC 2020 UHD Phase 3 8K, 4K HDR, 120Hz Rec 2020 10 bit / 12 bit /14 bit (1988) (1990) (1996) (1953) (2012) (2020)
- 28. Brian Kim Display color gamut 28 Pointer’s DCP-P3 sRGB BT2020 CIE 1976 u’v’ chromaticity diagram CIE 1931 xy chromaticity diagram Dominant Wavelength Primary v’ u’ y x 0.523 0.451 0.33 0.64 611.8nm Red Rec. 709 / sRGB 0.563 0.125 0.6 0.3 548.0nm Green 0.158 0.175 0.06 0.15 464.5nm Blue 0.523 0.451 0.33 0.64 611.8nm Red Adobe RGB 0.576 0.076 0.71 0.21 533.9nm Green 0.158 0.175 0.06 0.15 477.8nm Blue 0.526 0.496 0.32 0.68 614.9nm Red DCI-P3 0.578 0.099 0.69 0.265 544.2nm Green 0.158 0.175 0.06 0.15 464.5nm Blue 0.528 0.477 0.33 0.67 615.2nm Red NTSC 1953 0.576 0.076 0.71 0.21 533.9nm Green 0.196 0.152 0.08 0.14 470.0nm Blue 0.51651 0.55649 0.29203 0.70792 630.0nm Red Rec. 2020 0.58674 0.05573 0.79652 0.17024 532.0nm Green 0.12558 0.15983 0.04588 0.13137 467.0nm Blue y x 0.329 0.3127 white(D65) 0.351 0.314 white(DCI)
- 29. Brian Kim Color coordinates of typical LEDs 29 Color purity of Green LED: The green LED has a lower color purity due to the non-zero spectral width of the LED emission and the strong curvature of the chromaticity diagram in the green region. GaN (525nm) GaN (505nm) GaN (498nm) GaN (450nm) AlInGaP (590nm) AlInGaP (605nm) AlInGaP (615nm) AlInGaP (626nm)
- 30. Brian Kim 30 1. Obtain specification from customer Optical specification luminance, color gamut, R/G/B/W tolerance, optical power distribution, etc. Electrical specification max. power/current, etc. Mechanical specification Reliability test conditions 2. System design Optics design, driver / power, LED selection, etc. 3. Prototyping & test 4. Prepare Golden samples & approval / feedback May need a couple of loops 5. Small scale production Procedure to select LED DCI-P3 w/ tolerance
- 32. Brian Kim Perception of color by the human eye Physical Physiological Psychological 32
- 33. Brian Kim Chromatic light 33 Radiance (Watts: W) o The total amount of energy that flows from the light source. Luminance (lumen: lm) o A measure of the amount of energy an observer perceives from a light source. o Example: Far Infrared Region Brightness o Subjective descriptor of light perception that is practically impossible to measure. Brightness vs. Luminance: Technically speaking, brightness is what we perceive, whereas luminance is what we measure.
- 34. Brian Kim Luminance vs. Brightness Luminance o Measured amount of light in [cd/m2] o Physical quantity -the light energy weighted by the spectral sensitivity function of the human visual system -independent of the luminance of the surrounding objects Brightness o Perceived amount of light (perceived luminance) o Psychophysical quantity -depends on luminance of the surround (lateral inhibition, contrast) 34 Brightness Luminance
- 35. Brian Kim Relationship between Radiance and Luminance 35 Luminance – how bright the surface will appear regardless of its color. [units: cd/m2] o Photometric quantity defined by the spectral luminous efficiency function o L ≈ 0.2126 R + 0.7152 G + 0.0722 B Luminance Light spectrum (radiance) Luminous efficiency function (weighting) = x Luminosity curve = Eye response curve = Luminous efficiency curve
- 36. Brian Kim Effect of colors on brightness? Which patch looks brighter ? o The red and pink patches look by far the brightest, while blue, green, and amber are less bright. Dimmest of all is the green- yellow patch, third from the left. o Lights of these colors will behave in the same way as these printed examples. Red and pink lights will always look much brighter to the eye than a green or yellow light of the same luminance ●Helmholtz–Kohlrausch effect o The impact of color saturation on perceived brightness; the more saturated the colors, the brighter they appear. o Certain colors (green and yellow) do not have significant effect, however, any hue of colored lights still seem brighter than white light that has the same luminance. seven differently colored patches against a grey background. (all patches are same lightness ) 36 The strength of the H-K effect depends upon color and the saturation level, but it can increase the perceived brightness.
- 37. Brian Kim Contrast effect & Weber’s law The simultaneous contrast effect o The perceived brightness depends on the contrast. o The perceived brightness of inner circle are different due to a different background intensity levels even they are identical. o As the background gets darker, the perceived brightness of the circles increases. Weber’s law o Human's ability to resolve two visual stimuli with different intensities, L and L+ ∆L, is determined by the ratio ∆L/L over a wide intensity range. o The higher L needs a larger ∆L for a target to be detected rather than a small L. - x,y axis can be either “Stimulus Intensity and Sensation Intensity” or “Luminance and Brightness”. 37 ∆L/L= constant (~0.01)
- 38. Brian Kim A real-world scenes vs. digital images Real-world scenes ≠ Digital images o The restricted color gamut and even more constrained luminance and contrast ranges o Although tremendous progress can be observed in recent years towards improving the quality of captured and displayed digital images and video, the reproduction of real world appearance is still a farfetched goal. High dynamic range imaging (HDRI) o manipulate all colors and brightness levels captured by a camera o offer brighter and more colorful than their digital reproductions, but also contain much higher contrast Dynamic Range: The ratio between the max. and min. measurable light intensity 38