2012
2012 Color, Brightness, Contrast, Smear Reduction and Latency 2 Stuart Nicholson Program Architect, VE
Overview Topics Color Luminance (Brightness) Contrast Smear Latency Objective What is it? How do we quantify it? Why does it matter? 3
COLOR 4
What is color? Human eye is sensitive to EM radiation between 400-700 nm 3 types of cone structures (trichromacy) in the eye allow us to identify different spectral power distributions as different colors (Vanessaezekowitz @ en.wikipedia) 5
Describing Color CIE 1931 Color Space CIE 1931 XYZ mapped into x, y All visible colors represented by positive numbers Mixing two colors yields a color on line between the two colors 2D view does not show luminance CIE 1931 Y is close to photopic weighted luminosity L CIE 1931 is useful, but not perceptually uniform 6
Describing Color Color Gamuts ITU 709 argb Source1 Source2 Source3 We can mimic colors using mixtures of other colors (a metamer) that result in same cone response 3-primary display systems cannot reproduce all colors, typically saturated blue-green / cyan colors 7
Projected Colors and Metamerism Different people may perceive these color mimics as different colors This is called observer metamerism Typically due to slightly different cone responses (especially when mixing colors near highly sloped regions of cone response) Can be exacerbated if using narrowspectra colors with optics whose gain changes rapidly with observer position May cause changes cone response ratio that are perceived as different colors (Vanessaezekowitz @ en.wikipedia) 8
0.0 Normalized Intensity Cloudy Daylight - Blue srgb on LCD - Blue 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - Blue Sky srgb on LCD - Blue Sky 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - Green srgb on LCD - Green 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - Orange Yellow srgb on LCD - Orange Yellow 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - Red srgb on LCD - Red 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - White srgb on LCD - White 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - Purple srgb on LCD - Purple 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Normalized Intensity Cloudy Daylight - Yellow Green srgb on LCD - Yellow Green 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm]
Color Differences - Uniformity Ellipses illustrate regions of similar perceived color sensitivity for the average human Eye sensitivity to color variation (most to least) Blue White Red Green 10
Typical Color Gamuts Specified as CIE 1931 x, y times 1000 Name Red Green Blue EBU / PAL / SECAM 640, 330 290, 600 150, 60 SMPTE C / NTSC 1987 630, 340 310, 595 155, 70 ITU 709 / srgb 640, 330 300, 600 150, 60 argb (AdobeRGB 1998) 640, 330 210, 710 150, 60 11
100 Typical Color Gamuts CIE 1931 y * 1000 EBU / PAL / SECAM SMPTE C / NTSC 1987 ITU 709 / srgb argb(98) LED1 LED2 800 700 600 500 400 300 CIE 1931 y * 1000 750 730 710 690 670 650 630 610 590 570 550 150 170 190 210 230 250 270 290 310 330 350 CIE 1931 x * 1000 380 370 360 CIE 1931 y * 1000 90 80 70 60 50 40 30 20 10 200 340 330 100 320 310 300 0 290 0 100 200 300 400 500 600 700 800 280 CIE 1931 x * 1000 CIE 1931 y * 1000 350 600 610 620 630 640 650 660 670 680 690 700 CIE 1931 x * 1000 0 100 110 120 130 140 150 160 170 180 190 200 CIE 1931 x * 1000 12
LUMINANCE (BRIGHTNESS) 13
How do we talk about Brightness? Brightness is the term for how humans perceive light and is generally not quantified To provide a frame of reference, the average relationship between EM power and perceived brightness has been encoded in a luminosity function Thus we can convert radiometric power measurements into photometric measurements Luminance = EM spectrum weighted by a luminosity function Luminous flux = EM flux weighted by a luminosity function 14
What is Luminance? Human eye is sensitive to EM radiation between 400-700 nm Photopic: light adapted response Scotopic: dark adapted response Luminance is typically weighted for photopic response cd/m 2 (nits) or foot-lamberts Relative Sensitivity 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Scotopic Photopic CIE 1931 Y 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] Vos, J. J. (1978) Colorimetric and photometric properties of 2-deg fundamental observer. Color Research and Application Wyszecki & Stiles (1982) Table I (4.3.2) Color Science: concepts and methods, quantitative data and formulae (2 nd edition) 15
How do we talk about Brightness? There are many terms and many units Most common lumen for source brightness lux for light intensity incident on screen ft-l or nits for light intensity emitted from screen Perceived light intensity incident on surface Illuminance [lux=lm/m 2 ] also [ft-c = lm/ft 2 ] Perceived light intensity passing through or emitted from surface Luminance [cd/m 2 ] also [ft-l] Luminous Intensity [lm/sr] Luminous Intensity [lm/sr] Luminous Intensity [lm/sr] Luminous Flux or Luminous Power [lumen] = [lm] Total perceived light output from projector Projector For perfect Lambertian diffuse reflector 1 lumen of power incident on 1 m 2 would measure as 1 lux and reflect 1 cd/m 2 1 lumen of power incident on 1 ft 2 would measure as 1 ft-c and reflect 1 ft-l 16
White Point vs CCT Many colors have the same correlated color temperature (CCT) However, these colors are not all suitable as the white for a simulation application Std Illuminant 1.0 0.9 0.8 0.7 Scotopic Photopic CIE 1931 Y D50 345.67, 358.50 D55 332.42, 347.43 D65 312.71, 329.02 (Planckian-locus.png - Wikipedia) Relative Sensitivity 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength [nm] 17
Projector Native White Point Mismatches between the native white point of the projector and the white point of the gamut can result in reduced brightness Algorithms can be designed into single-chip projectors to mitigate this effect Brightness Relative to Native 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 3000 4000 5000 6000 7000 8000 9000 10000 White Point Temperature on Black Body Curve [K] Projector Native White Point (Black Body Temperature) 3500K 4000K 4500K 5000K 5500K 6000K 6500K 7000K 7500K 8000K 8500K 9000K 9500K 10000K 18
Brightness vs Color Gamut Helmholtz-Kohlrausch (HK) effect Images with saturated colors appear brighter Effect is stronger with saturated blues / reds compared to greens / yellows Brightness (perceived) is not the same as luminance (weighted physical power) 19
CONTRAST 20
What is contrast? Variation of brightness or color in an image For brightness, typically expressed as contrast ratio Sequential or full-on / full-off (FF or FOFO) Characterizes the maximum possible luminance difference a projector or display system can achieve over time (dynamic contrast) Checkerboard (one type is the 4x4 ANSI checkerboard) Characterizes the static contrast capability Susceptible to system configuration (cross-reflection, etc) Checkerboard square size affects measurement 21
Why does contrast matter? Good static contrast performance can reduce blending complexity for multi-projector systems Darker blacks reduces need for optical blend filters Good dynamic contrast (whether manual or automatic) can improve range of images that can be displayed without contouring System configuration and ambient light often dominate projector capability in determining final contrast performance 22
SMEAR 23
What is Smear? Screen Location Want to display object moving smoothly across screen Frame-based projection forces us to display object in discrete locations Unfortunately, due to smear object looks this big Want object to look this big Retina Location Screen Location Eye s field of view scanning across screen, tracking object Object moves across retina instead of staying in one location Object on screen Frame Frame Time Time 24
Smear Reduction Retina Location Reduced brightness Can flicker Reduced display time Higher update rate Much improved brightness and flicker Improved motion anti-aliasing Requires faster IG and video processors Screen Location Eye s field of view scanning across screen, tracking object Object moves across retina instead of staying in one location Reduced display time Reduced smear Higher update rate Reduced smear Object on screen Frame Frame Frame Time Time Time 25
L ATENCY 26
What is Latency? The difference between feeling like you re flying this... And this... (800px-Boeing_747-400LCF_Dreamlifter.jpg - Wikipedia) (f15-msf10-0135-003_med.jpg - Boeing) 27
Projector Latency Concept: delay from video input to image on screen But video takes a while to arrive, and... Displayed image takes a while to build up So, how do we define latency? Video Source Video Ingest Video Processing Imager Projection Lens Image on Screen 28
Latency IG Output Transport Latency Frame n Common definitions First-to-First T4 T0 Last-to-Last T5 T2 T1 = T0 = Video Cable Video Processing Modulated Light out Projection Lens First-to-First T4 T0 Projector receives first pixel from IG First pixel leaves projector video processor Frame n Frame n Frame n T0 T1 T2 T3 T4 T5 Video Processing Latency T4 = Imaging Latency Last-to-Last T5-T2 T5 = Last light of new projected image (image is fully formed) First light of new projected image T2 = Projector receives last pixel from IG T3 = Last pixel leaves projector video processor 29
Latency Smear reduction (reduction of hold or display time) can reduce latency Increasing update rate can significantly reduce latency IG Output Frame n IG Output Frame n Transport Latency Transport Latency Video Cable Video Processing Modulated Light out Projection Lens Frame n Frame n t0 t1 t2 t3 t4 t5 Video Processing Latency Frame n Imaging Latency Video Cable Video Processing Modulated Light out Projection Lens Frame n t0 t1 Frame n Frame n t2 t4 t3 Video Processing Latency t5 Imaging Latency 30
SUMMARY 31
Summary Color CIE 1931 x, y or CIE 1976 u, v Luminance (Brightness) White point Native white point cd/m 2 or ft-l Luminous power: lumens Contrast Static Dynamic Smear Hold time Update rate Latency Last-to-last 32
2012 Thank you. Questions?
2012 Thank you. Questions?