Karlheinz Blankenbach

Transcription

1 Karlheinz Blankenbach Pforzheim University, Germany Master in Embedded Systems, Pforzheim, SS 2008 Prof. Dr. Karlheinz Blankenbach Pforzheim University Tiefenbronner Str. 65 D Pforzheim, Germany Phone : Fax : kb@displaylabor.de Web : 1

2 Objective of Lecture Understanding of applications of electronic displays Know how of advanced display technologies Knowledge of relationships between electronics interface display Capability to design embedded (information) systems with the best display 2

3 Recommended Textbooks - L.W. MacDonald, A.C. Lowe: Display Systems, WILEY SID - P.A. Keller: Electronic Display Measurement, WILEY SID - Green, MacDonald: Color Engineering, WILEY SID - Ernst Lueder: LCDs : Addressing Schemes and e-o Effects, WILEY SID - Willem de Boer: AM Liquid Crystal Displays, NEWNES - R.L. Myers: Display Interfaces, WILEY SID - G. Berbecel: Digital Image Display, WILEY SID 3

4 Overview Introduction - Market overview - Some basics of metrology Selected Topics on Advanced LCDs, OLEDs and PDPs 3D & E-Paper (flexible) Display Technologies Touch Screen Technologies Display Interfaces 4

5 Worldwide Flat Panel Market Bill. $ CRTs - 50% of FPD in PC and TV only in developing countries AM LCD Source : isupply 10 PDP PM LCD 0 Other FPDs

6 Top 7 FPD Applications Worldwide Turnover / B$ ~ 100% AM LCD except for TV Notebooks Monitors Mobile Automotive TV Industrial Photo 6

7 Worldwide Flat Panel Market by Regions 7

8 Worldwide Flat Panel Market by Applications 8

9 Worldwide Units by Applications 2007: Production 60,000 m² 9

10 Shift of FPD Production From Japan to Korea & Taiwan to (?) China 10

11 AM LCD Application Waves 11

12 Worldwide Flat Panel Market : Small / Medium LCDs (< 10 ) Price per display shrinks! 12

13 Worldwide Flat Panel Market : AMLCD > 10 Mio. / B$ p.a SAMSUNG LG:PHILIPS AUO CMO SHARP CPT HANNSTAR INNOLUX BOE-OT Others Ranked acc. turn-over: compare e.g. SHARP and CPT 13

14 Worldwide Flat Panel Market : AMLCD > 10 Mio. / Units p.a SAMSUNG LG:PHILIPS AUO CMO SHARP CPT HANNSTAR INNOLUX BOE-OT Others Ranked acc. turn-over: compare e.g. SHARP and CPT 14

15 Worldwide Flat Panel Market : LED Backlight for LCDs LCD < 5 : ~ 100% LED 15

16 Worldwide Flat Panel Market : OLEDs Left: Linear rise or S-curve compare to mobile devices (right) Left: 10 M$ ticks reasonable? OLEDS in total are about 1.5% of total LCD-market in

17 Worldwide Flat Panel Market : Car Navigation 17

18 Worldwide Market : Head - Up Displays 18

19 Worldwide Market : Touch Screens 19

20 Worldwide Market : Low Power and Zero Power Displays 20

21 LCD TV Race (Prototypes) 2004 : : 82 Commodity up to 65 (2007) 2006 : 108 Gen 10 21

22 The Race for Size is limited by Mother Glass Size PANASONIC 150 PDP CES 2008 SHARP 108 LCD prototype Commercially in MP available: - LCD up to 70 - PDP up to 70 (PANASONIC 103 : 100,000 $) Forced by trend to signage (POI, small sized indoor billboards, 22

23 The Race for Size (I) Production issue Micrometer precision required in meter range! 23

24 The Race for Size (II) SAMSUNG Gen 8 = 1 B$ SHARP announced Gen 10 for

25 Flat Panel Display World Resolution QXGA HDTV SXGA XGA SDTV VGA QVGA P r o j e c t i o n Car MP, PDA 150 ppi p-si a-si Notebook PCmonitor LCD TV PDP Projection system Low information content displays Display Size / 25

26 How to Select a Display User quality Technical specification 'Good readability Optics, electronics, application, Magic Circle Pixel driving, e-o characteristics, rise & fall time, ghosting, Electro-optics Optics Luminance, contrast, grey scale, colour, response time, uniformity viewing angle, reflections,... Size, weight, price, lifetime, Driving, voltages, signal processing, power consumption, EMI, data input,... Electronics Application power supply, reliability, mature technology, temperature, vibration, displayed data,... 26

27 Overview Introduction - Motivation - Some basics of vision Basic Parameters (Merits) Common Issues Display Technology Dependent Issues (Shortcomings) Summary 27

28 Why Measuring Electronic Displays? Merits of display metrology Human vision is only descriptive Standardized measurement setups and test patterns Specifications enable judging of displays Wide range of measurement procedures for many applications Shortcomings of display metrology Ambient light (simulation) difficult, therefore most of the specified values are measured under dark room conditions Vision sees things that measurement can t capture and vice versa 28

29 Basic Characteristics of Vision Resolution Spectral Sensitivity for Day and Night Angular resolution : 1 at optical axis Example: 1m Time resolution ~ 50 ms (20 Hz) up to 100 Hz flicker sensitive ~ 500,000 colors can be distinguished plus signal processing (brain) 29

30 Sea of Measurement Tasks Environment Temperature, % rh, EMI, power supply, Driving system Signal processing Test patterns DUT Full, box, checkerboard, grille, grey level, color, dynamic, moving, Viewing angle Dark room or (simulated) ambient light Procedure, data storage Measurement of luminance and/or color Spatial Temporal - Jitter - Warm up - T-change - Lifetime 30

31 Overview Introduction Basic Parameters (Merits) Color Calculator: - B/W, GS, color - L, C R, GS, color - Parameters & vision Common Issues Display Technology Dependent Issues (Shortcomings) Summary 31

32 Photometric Units Units: RI Color Calculator Luminous flux (power) : F / lm Luminance cd : L / 2 m (emitter) Illuminance lm : E / lx = 2 m (receiver) Illuminance = power / area : df E = = da F A Examples - Luminous flux F = 1,000 ANSI lm for projector - Luminance L = 300 cd/m² for AMLCD - Illuminance for 1,000 lm projector at 2 x 3 m² E = 167 lx compare to 500 lx recommended at workplace! 32

33 Luminance [L] = cd/m² Typical values : Displays 50 1,000 cd/m², bulb 10,000 cd/m², sun at noon 10 8 cd/m², full moon 100 cd/m² Basic value for contrast, grey scale, uniformity, viewing angle,... One of three parameters for color measurements (Tristimulus) Measurement conditions - Dark room unless otherwise noted - Centre of display, perpendicular incidence - Usually F.O.V. of > 25 pixel for monitors - V(λ) corrected devices Same for color 33

34 Example of Optical Specification for AM LCD (II) Recommendation: Use same device as in spec! 34

35 Contrast Ratio is not a measure for readability of text (full screen)! Luminance ratio of bright / white / max / ON to dark / black / min / OFF Various definitions, mostly contrast ratio used, MTF s. b. Contrast ratio C L bright R = = L dark L L white black... Example Paper 10 : 1 L white L black Remarks High C R can be critical because of measurement error for L black Vision range : C R = 3 : : 1 C R 10 : 1 recommended for non-fatigue reading (paper!) High C R can bother (e.g. car headlights at night)! C R in specs measured without ambient light! E C R Various conditions like full screen, checkerboard, 35

36 Contrast Ratio Test Patterns Full screen LCD, OLED, PDP C R 100:1 Used for PDP C R 1,000:1 Centered box Checkerboard (Projection) Intra-Character C R depends on test pattern, especially for - PM drive (ghosting) - PDP - Projection 36

37 LED LCD - Backlight Improvements : Local Dimming High Contrast L max = 630 cd/m 2 L min = 0.03 cd/m 2 Power saving 50% average (image dependant) C R = 630 / 0.03 = 21,000 : 1 37

38 Modulation Transfer Function & Resolution Contrast modulation = MTF C Grille M = L L bright bright + L L dark dark Measure for text readability Threshold - Text C Grille M 0.5 n - Images C Grille M 0.25 Resolution Number of R = n T,I lines (Analogue) bandwidth Ghosting, 38

39 Grey Scale Measurements Greyscale for images, Results: Gamma, GS resolution, Gamma is seldom specified rel. Luminance LCD 18" LCD 8.4" L ~ D 2.3 CRT rel. Grey Level log (norm. Luminance) Grey bars or full screen log(luminance) gamma 2.5 log(luminance - L(0)) Contrast ratio only max / min! -2.5 Deviations log (norm. Grey Level) 39

40 Color Space CIE co-ordinate : Luminance L = Y Oldest CIE standard Still in use today but CIE 1976 UCS recommended in display metrology specs! Problem: Co-ordinate differences color differences (Mac Adam) Linear transformation of Tristimulus values, L = Y 40

41 LCD CIE 1931 Specification Plot these tolerances in CIE 1931 with x and y = 0.1! 41

42 Color Space CIE 1976 UCS 3. co-ordinate : Luminance L = Y CIE 1976 UCS recommended in display metrology specs! v Co-ordinate differences colour differences Linear transformation Gamut (100% usually refers to NTSC) u 42

43 TV Color Spaces YIQ : NTSC Y I Q = R G B YUV : PAL, SECAM Y U V = = = 0.30 R G (B Y) (R Y) B YCbCr : ITU-R BT.601, 709 HDTV Y Cb Cr = R G B All parameters are normalized Y Luminance, same formula for all TV color spaces IQ, UV, CbCr Chrominance 43

44 Color Space Summary CIE 1931 CIE 1976 UCS CIE Lab CIE Luv RGB Tristimulus transformation Linear Linear Non-linear n. a. Luminance Yes Yes No n. a. Plot 2D 2D 3D 3D Pros history co-cord. color JNDs SW Cons co-cord. color (Mac Adams) L not brightness 3D, L* L Device specific Used for Specs Specs (recomm.) E SW 44

45 Overview Introduction Basic Parameters (Merits) Common Issues - Spatial domain: Uniformity, MTF, gamma, color tracking, - Time domain: - Short: Flicker, jitter, swim, blur, - Medium: Warm up, - Long: Lifetime, - Ambient light: C R, GS, color, - Signal processing Display Technology Dependent Issues (Shortcomings) Summary Here: Some examples 45

46 Uniformity Deviation of a display parameter, e.g. luminance within display area 5, 9 or 13 spots method (ISO 9241) Definitions L ± L L L max min x % max L max < L min x % < Examples* 190 cd/m² ± 5% x % 10% 111% *:L max = 200 cd/m², L min = 180 cd/m² Area (linear or false color) Uniformity versus Campbell-Robson 46

47 Examples of Uniformity Measurement in Spec Spot size acc. FOV and distance Luminance variations outside this points are not specified like degradations due to burn-in or image sticking! 47

48 Color Tracking v' Color Tracking 20" LCD, CIE 1976 UCS grey white Color coordinate of white depends here on grey level! γ γ γ! u' 48

49 Contrast Ratio with Ambient Light Diffuse C R = L L white black + ambient light C Components of reflections R = L L white black (0 lx) + L (0 lx) + L reflected reflected L white L L white >> L reflected L reflected >> L black (0 lx) reflected L diffuse reflected = π L r E π (0 lx) r E white LCD (left) : specular & haze OLED (right) : mainly specular (diffuse with low intensity) 49

50 ISO Sunlight Simulation Diffuse or specular geometry Contrast Ratio Diffuse Geometry 1, LCD Measured C R Fitted C R ,000 40,000 60,000 Illuminance /lx 50

51 Examples for Contrast Ratio Degradation L White L Black C R (0 lx) C R with L Reflected = 10 cd/m² * / 11 = / 11 = / 15 = / 15 = 7 * : L Refl. (W) = L Refl. (K) Factor 5 for white luminance, L(K) = const. factor ~ 4.7 in C R for ambient light Factor 5 for black luminance, L(W) = const. factor ~1.4 in C R for ambient light for L White = 100 cd/m² and L White = 500 cd/m² Luminance of white has more influence on C R with ambient light than luminance of black and C R under dark room conditions! 51

52 Ambient Light & Grey Scale log (Luminance) Ambient Light 12 Transmissive LCD 33 % 0 lx 50 lx 200 lx 1,000 lx 2,000 lx γ = log (Grey level) Ambient light reduces the number of distinguishable grey shades (JND) Grey shades below 80 (of 255) can t be resolved for 2,000 lx 52

53 Ambient Light & Color y CIE 1931 & Ambient Light Reflections shift the color coordinates towards the locus of the light source x 53

54 Display Technology vs. Ambient Light : Sunlight Outdoor Display is ON! E-Paper is close to paper Reflective b/w PM LCD Transmissive color AM LCD Transflective Color AM LCD Reflective E-Paper Reflective MUX LCD 54

55 Blur Effects Spatial Blur by limited bandwidth, jitter, e.g. MTF, analogue path see Display Metrology Luminance Blur T Frame t Motion Blur by hold type displays (e.g. AM) Luminance Hold Type T Frame t 55

56 Motion Blur Basics Motion blur is caused by AM techniques due to lack of auto-tracking by human vision (PDP has similar issues on subframe coding) Impulsive displays like CRTs don t suffer of motion blur Adapt AM to impulsive drive Visualization Perceived AM CRT Display & hold Flashing 56

57 Display & Observer Motion Blur Blurry edges! 57

58 Motion Blur Reduction Techniques Only with LED backlight! 58

59 Displaying Moving Objects Original Non - Hold (2 pictures movie ) Type (CRT) Hold Type (AM technologies) Frame 1 (Picture 1 is completely written on the display) Frame 2 ( T = Frame 1 + 1/2 Frame 2) (Half of picture 2 is written on the display) Leftover of frame 1 not visible due to phosphor decay Smearing! ¾ T Frame

60 Effect of Eye Movement on CRTs (Impulsive Drive) White box is moved with 4 pixels per 60 Hz frame Examined line Display x On Display, x CRT On eye, x 1 F 1 F t Eye movement t Perceived image by eye 60

61 Effect of Eye Movement on 60 Hz LCDs (Hold Type) On Display, x 60 Hz LCD On eye, x 1 F 1 F t Eye movement t Perceived image by eye Unsharp edge blurring of 3 pixels 61

62 Effect of Eye Movement on 120 Hz LCDs (Hold Type) On Display, x 120 Hz LCD On eye, x 1 F 1 F t Eye movement t Perceived image by eye Unsharp edge reduced blurring of 1 pixel 62

63 Effect of Eye Movement on Displays White box is moved with 4 pixels per 60 Hz frame Examined line Display x Perceived image by eye CRT 60 Hz LCD 120 Hz LCD - CRT has best motion picture quality - LCD with 120 Hz is better than 60 Hz 63

64 Black Frame Insertion Subjective Evaluation for Moving Picture L L peak BFI > L Hold BFI CRT Hold 1 F t (not to scale) ITU Grade (International Telecommunication Union) Excellent Good Fair Poor Bad Black frame insertion CRT OCB w. LED TN Scroll velocity /pixel per frame OLED : Easy Fast OLED response Only SW required LCD High LC switching speed required Black frame insertion or impulsive LED backlight

65 Impulsive Backlight for LCD Hold - Type Technologies Frame 1 Frame 2 Write grey level data to display Write grey level data to display No smearing Backlight OFF ON OFF ON T Blackframe T Display 8.3 ms instead of 16.7 ms = ½ T Frame = ½ T Frame fast switching LC required Similar: Scrolling backlight with LEDs

66 Impulsive Backlight for LCD Hold - Type Technologies Could be also made by 120Hz drive with GL = 0 insertion. Application: Rolling backlight / black frame 66

67 Impulsive Backlight for LCD Hold - Type Technologies No BFI With BFI 67

68 Impulsive Backlight for LCD Hold - Type Technologies ON OFF Scrolling of black backlight sections 68

69 Motion Blur Reduction by Higher Frame Rate Motion blur reduction using high frame rate driving: 120Hz/100Hz Requires new panel design, also video sources low frame rates (NTSC = 60Hz, PAL = 50Hz, Film = 24Hz) Motion interpolated frame technology or Black Frame Insertion needed such as McFi (Motion Compensated Frame Interpolation)

70 Comparison by SAMSUNG BFI

71 Comparison by SIEMENS M. Zachmann (Pforzheim U) et al. BFI BFI Evaluate your display with PIXPERAN (freeware)

72 CLEAR MOTION DRIVE by JVC 100 or 120 Hz & interpolated images

73 Motion Blur Reduction Using 120 Hz & 72 Hz Driving with Motion - Compensated Frame Interpolation NTSC Video 60 Hz 120 Hz calculated 120 Hz cheaper than backlight methods Movie 24 Hz 72 Hz (DVD) For LCD and PDP Motion judder for 3:2 pull-down!

74 Frame Rate Conversion Techniques (I) PAL Video 50 Hz Motion blur on Hold type displays (LCD, ) PAL Video 100 Hz repeated judder PAL Video 100 Hz motion compensated no judder (if algorithm OK) calculated t

75 Frame Rate Conversion Techniques (II) Movie 24 Hz (24p) Original Movie 60 Hz 3:2 pull down repeated judder, jerkiness Movie 72 Hz 3:3 pull down repeated judder Movie 120 Hz motion compensated no judder (if algorithm OK) calculated t

76 PDPs : Motion Artefacts by Sub - Frame Driving Many motion artifacts etc. of PDPs can be minimized by massive signal processing

77 Overview Introduction Basic Parameters (Merits) Common Issues - Luminance domain: Loading, halation - Time domain: Burn-in, differential ageing, image sticking - Observer domain: Spatial & time invariant - Response time - Viewing angle: L, C R, GS, color Display Technology Dependent Issues (Shortcomings) Summary Here: Some examples 77

78 LCDs : Response Time Effects Rel. L T ambient = 25 C Grey level transition T Rise /ms T Fall /ms T > frame time (16.7 ms) T Rise > T Fall T = T(grey 1 grey 2) Grey Level t /ms Grey level & color shifts 78

79 LCD Motion Blur caused by Slow Response Time Rise time often fall time Fall time Rise time Moving Picture Response Time (MPRT) by Brightness Edge Width Measurement setup: Camera and screen (one moving, one fixed) Luminance A - A' 90% A Response time depends on grey level (start, final) Captured image A' Motion blur occurs for all AM technologies due to vision (Hold type) 10% BEW Pixel coordinate 79

80 Motion Blur Measurement Procedures t HS Capturing so that image stands still on the camera (chip) High speed camera High speed sensor, problem: t x 80

81 MPRT & Perceived Motion Blur (I) 81

82 MPRT & Perceived Motion Blur (II) 82

83 Motion Picture Response Time by Brightness Edge Width Normalising : N-BEW = BEW / moving speed MPRT = N-BEW (G2G) 83

84 Motion Picture Response Time by Brightness Edge Width MPRT = 16 ms 84

85 Motion Picture Response Time by Brightness Edge Width MPRT 85

86 MPRT Measurements by Grey - to - Grey Response Time CIE 1976 UCS 6 equidistant brightness e.g. L = 0, 2.2, 7.7, 18.4, 36.2, 62.8, 100 cd/m2 86

87 MPRT Measurements by Grey - to - Grey Response Time Extended Blurred Edge Time 87

88 Viewing Angle L, C R, Clear LEDs Lambertian (paper) Emissive displays, diffuse LEDs, standard screen LCDs, high gain-screen Angle Angle isn t everything! Measurements : 2D, 3D, value : ± or Σ 88

89 Viewing Angle Basics PM LCD 6:00 AM LCD seen from 12:00 displaying green 6:00 z θ 9 / 12 y Φ 180 / 9 Dis play x 270 / 6 0 / 3 2D Display Sensor 3D 89

90 Contrast Ratio Dependency from Viewing Angle for LCD 160 C R " LCD 0lx, horizontal 80 Threshold definition! Viewing Angle / Often C R min > 5 : 1 Specify viewing angle only with the minimum contrast ratio (or better color difference E) here : C R > 100 : 1 ; C R > 50 : 1 ; C R > 20 : 1 C R > 5 : 1 90

91 Example from LCD - Specification Threshold definition 91

92 Iso - Contrast Plots of AMLCDs A AMLCD B cm Scale for contrast ratio Display A is better because of Larger area with nearly constant contrast ratio for perpendicular incidence Larger maximum contrast ratio In both cases a 6:00 observer position with θ 15 is recommended 92

93 Grey Scale Inversion L Visual test pattern L = L dark grey - L black Grey scale inversion θ GI θ θ GI θ Remarks - If the sign doesn t change and luminance levels only merge, the image quality is horrible, but grey scale inversion doesn t occur in meaning of standard (norm) - Contrast inversion is the extreme case of grey scale inversion, when maximum and minimum luminance levels change 93

94 Contrast Ratio vs. Grey Scale Performance C R = 200 : 1 L /cd/m² 10 Φ = 0 / θ = 0 C R = 400 : 1 C R = 200 : 1 C R = 400 : Grey level (8-bit) Φ = 90 / θ = 20 Contrast Ratio values are misleading here because a reduction of only 2 occurs but the grey level representation is completely vanished (flat curve, blue) compared to perpendicular incidence (OK, brown) no grey shades visible for 20 off! 94

95 Color Dependency from Viewing Angle of LCDs Measured color representation, black was shown on the screen! Inversion! Visual test pattern 95

96 Colour Wash Out Electro-optical curve changes with viewing angle! 96

97 Luminance Domain : Loading of Plasma Panels Luminance of centered white box depends on their size, this is called loading. Why? Power limitation This can be also observed for CRTs (incl. image size variations). 97

98 Aging of Displays Lifetime (50%) Full screen white measured Note: Life time measured at ~ 75 C, and extrapolated to 25 C in specs! Temperature Life time Differential Aging Different lifetimes of color subpixel color shift Static content like on airport information displays can cause burn-in and image sticking L L rel. L 1 Inital drop No inital drop T LT T LT t t 98

99 Burn - In of Emissive Displays and LCD Image Sticking CRT PDP Several hours display dark grey LCD 99

100 Burn - In and LCD Image Sticking Measurement - T = 0 : Display 5 x 5 (black & white) checkerboard and mark 3 boxes (2 white, 1 black) - Measure initial luminance for the 3 boxes for black and white: L WL, L WC, L BL, L BR - Burn-in display when showing the checkerboard, if applicable with high temperature for speeding up the effect, e.g. for OLEDs - Wait x hours - Measure luminance of the same locations for full white (K WL, K WC, K WR ) - Measure luminance of the same locations for full black (K BL, K BC, K BR ) - Calculate residual image factors by Simpler test: L WL(x) + LWR(x) Burn in (%) = LWC(x) - Checkerboard as burn-in - White screen as test pattern after x hours IS: Test at grey (60/255) R s are greater than 1 because of aged luminance K is lower than initial luminance L 100

101 Overview Introduction Basic Parameters (Merits) Common Issues Display Technology Dependent Issues (Shortcomings) Summary 101

102 Parameters Affecting Measurements & Visual Perception - Photometric - Colorimetric - Resolution - Temporal ~ - Directional variations - Lateral ~ Signal processing 102

103 Example of Optical Specification for AM LCD Inverter 103

104 Overview of Standards Parameter Standard Display Parameter Standard Display Luminance CIE 18,15.2 IEC 47(CO)16 IEC 61747* CECC A3 Chromaticity IEC 47(CO)16 ISO 9241/3 ISO IEC 61747* ISO 9241/3 Luminance and ISO chromaticity uniformity CECC WG 20B IEC 61747* ISO Response time CECC A3 IEC 47(CO)16 IEC 61747* All LCD LCD LCD CRT FPD CRT FPD LCD FPD LCD LCD Reflections ISO CECC A3 IEC 47(CO)16 ISO 9241/7 FPD LCD LCD CRT IEC 61747* Electro-optical Transfer EBUTech 3273 CRT, LCD Resolution CECC11000 all Crosstalk CECC WG 20B ISO LCD FPD IEC 61747* Distortion ISO 9241/3 CRT Viewing direction CECC A3 IEC 47(CO)16 ISO LCD LCD FPD * : for LCD matrix displays Common: VESA FPDM, SPWG, TCO 104