Background Statement for SEMI Draft Document 4571J NEW STANDARD: TEST METHOD OF PDP TONE AND COLOR REPRODUCTION

Transcription

1 Background Statement for SEMI Draft Document 457J NEW STANDARD: TEST METHOD OF PDP TONE AND COLOR REPRODUCTION Notice: This background statement is not part of the balloted item. It is provided solely to assist the recipient in reaching an informed decision based on the rationale of the activity that preceded the creation of this Document. Notice: Recipients of this Document are invited to submit, with their comments, notification of any relevant patented technology or copyrighted items of which they are aware and to provide supporting documentation. In this context, patented technology is defined as technology for which a patent has issued or has been applied for. In the latter case, only publicly available information on the contents of the patent application is to be provided. Background Different types of flat panel displays (e.g., LCD, PDP, and OLED) have different features and characteristics depending on their nature like luminance by the APL characteristics. For example, the maximum luminance of PDP varies according to the average picture level (APL) by its limitation of power consumption characteristics. The APL characteristics are not always disadvantages but sometimes advantages from the perspective of eye fatigue while watching TV. So, PDP module and PDP TV need a reasonable standard for display characteristics. In order to determine a reasonable standard, we can consider a fixed window measurement, for example /6 of the display (/4 of the display both horizontally and vertically) to test gamma. For most scenes the average brightness does not change suddenly; the variation of APL during watching TV has almost no influence on the picture quality. This document is necessary to set an industry standard on PDP tone and color reproduction (gamma and color gamut accuracy) for estimation of basic picture quality for PDPs. Revision Control Date Version Name Edits June, Kyungjin Kang First version of this document October, 008. Kyungjin Kang February, 009. Kyungjin Kang Improved English Document was transferred to the FPD Metrology Committee to resolve scope issue. January, 00.3 Kyungjin Kang Reflected reject comments, Improved figures February, 0.4 Kyungjin Kang Failed to get 60% return rate June, 0.5 Kyungjin Kang Reflected reject comments, Improved figures January, 0.6 Kyungjin Kang Reflected reject comments, Improved figures March, 0.7 Kyungjin Kang Reflected reject comments, Improved figures February, 03.8 Kyungjin Kang Reflected reject comments, Improved figures June, 03.9 Kyungjin Kang Reflected reject comments, Improved figures Review and Adjudication Information Task Force Adjudication Group: To be reviewed at the Korea Tone and Color TF meeting Committee Adjudication Korea FPD Metrology Technical Committee

2 Date: Thursday, November 7, 03 Thursday, November 7, 03 Time & Time Zone: 3:00 5:00 (KST) 5:00 7:00 (KST) Location: SEMI Korea office SEMI Korea office City, State/Country: Seoul, Korea Seoul, Korea Leader(s): Document Originator(s) Standards Staff: Kyung Jin Kang (LG Electronics), Kyung Jin Kang (LG Electronics), Natalie Shim (SEMI Korea) Jong Seo Lee (Samsung Electronics) Il-Ho Kim (LMS) Kyung Jin Kang (LG Electronics), Natalie Shim (SEMI Korea)

3 SEMI Draft Document 457J NEW STANDARD: TEST METHOD OF PDP TONE AND COLOR REPRODUCTION Purpose. This test method describes the measurement methods of principal characteristics of PDP (Plasma Display Panel) or PDP-TV, which is related to TV standard reproduction characteristics of fidelity to original video.. This standard is used to define methodologies used for measurement of PDP s native features based on APL (Average Picture Level). Scope. This standard is applicable to display quality evaluations standard of PDP modules or PDP TVs, but the metrology except APL luminance related equations, it can be applied to the other display device like LCD and OLED.. The standard includes practical methodologies of tone and color reproduction measurements. It does not address optical performance like contrast or brightness..3 Tone and color of display device is basic features of a general display device. For standardization of PDP, high accuracy and fidelity to these standard features are necessary..4 This standard is only for general purpose for display methodology. Hence special languages for special features or phenomena are not included. NOTICE: SEMI Standards and Safety Guidelines do not purport to address all safety issues associated with their use. It is the responsibility of the users of the Documents to establish appropriate safety and health practices, and determine the applicability of regulatory or other limitations prior to use. 3 Referenced Standards and Documents 3. CIE Standards 3.. CIE Standard colorimetric system 3.. CIE 93 xy coordinates color space 3..3 CIE 976 u v coordinates uniform color space 3. ICDM Standards 3.. IDMS Information Display Measurements Standard, Version.03, June ITU Standards ITU-R BT Parameter values for the HDTV standards for production and international program exchange, SMPTE Standards SMPTE 303M-00 Color Reference Pattern, August Referenced Documents 3.5. R.W. G. Hunt, Measuring Color, nd ed. Chichester, U.K.: Ellis Horwood, 987. NOTICE: Unless otherwise indicated, all documents cited shall be the latest published versions. CIE International Commission on Illumination, Kegelgasse 7, A-030 Vienna, Austria International Committee for Display Metrology, International Telecommunication Union Telecommunication Standardization Sector, 4 The Society of Motion Picture and Television Engineers, Page 3 Doc. 457J SEMI

4 4 Terminology 4. The definitions of general terms follow IDMS.03 and its following version. 4. Acronyms and Abbreviations 4.. APL average picture level 4.. CCT correlated color temperature 4..3 PDP plasma display panel 4.3 Optical Terminology 4.3. tone the gradation of output brightness by a display input variation. This usually means the gradation of gray, but color tone also includes the gradation of color color the perceptual concept from hue and saturation of video signals including chrominance components. Generally the term color might be defined to include black, white and gray colors that have no chrominance component, but here color is defined as output with some chrominance components IRE an arbitrary unit used to describe a video signal. White is defined to be 00 IRE and the blanking level is defined to be 0 IRE. It stands for Institute of Radio Engineers. 4.4 Measurement Terminology 4.4. average picture level (APL) there are two kinds of APL, pre-gamma APL (Type APL) for gamma corrected input signal(r, G, B) and post-gamma APL (Type APL) for gamma de-corrected panel display signal(r, G, B ). APL without any notice usually indicates the post-gamma APL because APL is used for panel load processing. The post-gamma APL means the average of the total R, G and B data after gamma de-correction, not input signal (gamma corrected) values. The unit of APL is usually % and generally just APL means the post-gamma APL. The APL of input signal is greater than the post-gamma APL. The gamma de-corrected value of the input signal average is less than or equal to APL, where the equal case is only when all input data are equal or 0/00 IRE data correlated color temperature (CCT) a term used to describe the color of a light source whose chromaticity lies close to the Planckian (black body) locus on a CIE 960 chromaticity chart. In general, it is the temperature of a black body radiator which is the most similar to chromaticity of the evaluated light source and it is expressed in degree Kelvin. CCT is used to adjust white balance gamma power relationship between R, G, or B input signals and their output luminance. It is applied to subdivide low level s tone resolution with limited number of bits, and expressed by form(v )γ of power function and it is specially called decoding gamma (γ_d). Conversely, its reciprocal is called encoding gamma (γ_e) which is processed on video camera side. Normally, decoding gamma(γ_d) = primary color one of a set of colors that are combined to produce any desired set of intermediate colors, within a limitation called the gamut. The primary colors for color television are red, green, and blue. The exact red, green, blue colors used are dependent on the television standard. Display devices do not usually use the same primary colors, resulting in minor color changes from ideal white balance adjustment or accuracy of color coordinates at a white input level, usually equal to or greater than 70% APL of full screen gray balance adjustment or accuracy of color coordinates at all input ranges, usually from 5 IRE to 00 IRE inputs. It is explained by all range white balance meaning luminance uniformity In IDMS, it is calculated by Minimum luminance divided by Maximum luminance among the defined screen positions and presented by %. For the measurement positions of a screen, refer to IDMS 5 Test Setup 5. Setup conditions 5.. All measurements shall be carried out in a dark room normally under lux. Page 4 Doc. 457J SEMI

5 5.. Warm up time of the display is at least an hour for all measurements, if not specified by the manufacturer of the display For tone measurement, all picture quality enhancement functions shall be bypassed to preserve the native picture qualities (luminance curve, APL, etc.). When some picture quality enhancement functions are active in the measurement, it shall be recorded in the report. When the measurement is applied to some picture mode of a PDP set (normal, economy, dynamic, living room, etc.), it shall also be recorded in the report. And any picture mode should be at just-scan mode without over-scan of image. 5. Measurement Setup 5.. The measuring equipment shall be set as shown in Figure for non-contact LMD (light measurement device). The optical axis of non-contact LMD shall be normal to the centre of the display surface. The measuring point is the center of effective screen area of the display. When other point is measured, it shall be noted in the report. 5.. During the measuring process, any residual level or image sticking by previous images should not exist on the display screen. The measurement order of a tone pattern shall be from lower level to higher level as preventing residual level as possible. Plasma display Light measurement device Signal generator Measurement point Control PC Figure The Measurement Set-up 6 Tone Measurement 6. Each test pattern to be displayed may occupy some percentage of area at over-scan screen mode, but it is not concerned whether it is at just-scan or over-scan mode. To match the results in PDP module and TV set, however, the just-scan mode shows the accurate results of PDP module and TV set. 6. Constant APL Gamma 6.. Gray-bar Window Gamma Page 5 Doc. 457J SEMI

6 6... Since PDP luminance for inputs varies by APL, the gamma values are measured at 0% to 90% APL(postgamma APL) separately with steps gray-bar pattern varying each input data step by 0%. In this document, just APL means the post-gamma APL When it is measured on a screen, the background signal level is changed for each APL while maintaining gray-bar pattern at the screen center For the measurement of avoiding luminance non-uniformity, each box of luminance levels is moved to center of measurement point of Figure where change the positions each other in regular sequence. NOTE : Example of area ratio : Each bar of gray pattern : 0.98% NOTE : The whole gray-bar pattern : 0.97% x = 0.78% NOTE 3: Background area : 00% % = 89.% Figure General Pattern of Gray-bar Window Gamma The APL of the pattern in Figure can be calculated by display module inputted values after gamma decorrection in range 0~55 of 8 bits. In case of conventional. gamma (shortly. gamma), APL(%) = (/55). () In Equation (), 0.33 means the APL of step gray-bar data. Outer background data is calculated using the following equation. = 55 ( ( APL ) / 89. ) /. () For ITU-R BT. 709 (shortly BT. 709) gamma, the APL and value are also calculated by a similar method but it is substituted by. gamma equation because APL is just the approximate test case Figure 3 shows an example of a gamma measurement graph in 9 kinds of output APL(post-gamma APL). Page 6 Doc. 457J SEMI

7 Output Luminance (cd/m ) APL 0 APL 30 APL 40 APL 50 APL 60 APL 70 APL 80 APL 90 APL 0 IRE 0 IRE 0 IRE 30 IRE 40 IRE 50 IRE 60 IRE 70 IRE 80 IRE 90 IRE 00 IRE Input Signal Level Figure 3 An Example of Input-Luminance Curve by APLs 6.. Single Window Gamma 6... Single window gamma measures white luminance according to APL change of Plasma module in a fixed window In order to fix the APL, luminance of outer area would be changed down from high to low IRE while a center box is measured from 0 to 00 IRE which means 0 to 55 gray levels The size of the box is /5 horizontal and vertical pixel numbers of source video, using approximately 4% of the total screen area in a display. 4% screen area Outer data (Keeping for an APL) Inner data X (Measured) Figure 4 General Pattern of Single Window Gamma Page 7 Doc. 457J SEMI

8 6...4 The APL of the pattern in Figure 4 is calculated by gamma de-correction in 8 bit. In case of conventional. gamma, APL(%) = 4 (X/55) (/55). (3) Outer background data is obtained using the following Equation (4). = 55 ( ( APL 4 ( X / 55 ). ) / 96 ) /. (4) 6..3 RGB Gamma The luminance gamma measurements in above are also used for RGB gamma measurement The explanation is abbreviated because it is a similar method except for APL calculation The inner window s APL with one color among R, G or B would be one third of Gray pattern. 6.3 Constant-luminance APL Gamma 6.3. Generally as PDP s characteristics depend on the APL, the luminance curve may be approximately characterized by an exponential curve as shown in the following Equation (5) and Figure 5. b L ref [ cd / m ] c (5) APL a NOTE 4: a, b, c integer constants to get higher accuracy for a PDP module. Typically, a is below 50 (about 0), b is above 5000, c is below 50. NOTE 5: APL average picture level in percentage, generally above 6% might be applied. Max. Lumi. % 5% 00% APL(%) Figure 5 An APL-Luminance Curve Modeling for PDP and CRT Sometimes this approximation may be not suitable at low APL (generally under 6% APL similar to CRT), because of the PDP s optical performance limitation due to the lack of the sustain period etc Sometimes, as suggested by the dotted line in Figure 5, a flat region is included at low APL for maintaining constant luminance When a PDP has the APL range with constant luminance up to over 4% APL, it is measured without varying background level to maintain an APL This measurement is made on the center screen while a box of varying gray levels is displayed. The box of a center is fixed up with a box of 0 to 00 IRE The size of the box is /5 to / horizontal and vertical screen size of the display, using 4% to 5% of the total screen area. Page 8 Doc. 457J SEMI

9 The larger window means that the display has a constant luminance curve from 0 to the higher APL as large as the window size In the previous APL luminance equation, to evaluate and compare the shape regulation of the APL luminance curve with CRT in 6% to 00% APL range, APL luminance regulation accuracy is defined as following Equations (6), (7), (8), and Figure 6. L _ Re gaccu. [%] 00 LAPL Lref 7 NOTE 6: L APL Maximum Luminance (cd/m ) at each APLs (4, 9, 6, 5, 36, 49, 64, 8, 00) White Luminance (cd/m ) APL PDP A Reference for PDP A PDP B Reference for PDP B (6) APL (%) Figure 6 Examples of Maximum White Luminance Curve by APLs In these examples, for PDP A, Equation (7) is adapted to get the regulation accuracy 93.5% Lmax[ cd / m ] (7) APL For PDP B, Equation (8) is used to get the regulation accuracy 96.4% Lmax[ cd / m ] 0 (8) APL Luminance Gamma Luminance gamma measures the luminance of gray patterns which change the gray value from white to black with the same linear steps in 4% to 5% boxes A level of outer area is fixed at 0 IRE and its APL is not constant. Usually this measurement is meaning at flat luminance range like the dotted line of above Figure All gray patterns are presented sequentially from 0 to 00 IRE in steps of 5% to 0%. Page 9 Doc. 457J SEMI

10 Output Luminance (cd/m ) 0 IRE /5H ~ /H/ 4 ~ 5% window 0~00 IRE /5V ~ /V Figure 7 Window Pattern of Gamma Measurement 0 0 IRE 0 IRE 0 IRE 30 IRE 40 IRE 50 IRE 60 IRE 70 IRE 80 IRE 90 IRE 00 IRE Input Signal Level Figure 8 An Example of Gamma Measurement Result RGB Gamma RGB gamma measures the luminance scale of the primary colors (R,G,B) using ~ color levels at 0% or 5% steps As the luminance gamma, 4% ~ 5% area window changes from black to highest level of the colors, and calculate a gamma value for each color. Page 0 Doc. 457J SEMI

11 0 IRE 0 IRE 0 IRE 4 ~ 5% window Red 4 ~ 5% window Green 4 ~ 5% window Blue Figure 9 Window Pattern of RGB Gamma Measurement 6.4 Gamma Value Calculation 6.4. Table is the data of standard gamma for about 5% step inputs In ITU-R BT.709, under 8.% of input data is not used because it is a linear slope region. So, gamma calculation for ITU-R BT.709 is done so that the luminance of 8.% input is treated as an offset. Table Gamma De-corrected Luminance by 8 bit Inputs Input Values, I % Full Scale (IRE) 8 bit Integer Inputs (0 55, I k ) ITU-R BT. 709 % Full Scale Display Luminance. Gamma % Full Scale (*) Page Doc. 457J SEMI

12 Input Values, I Display Luminance NOTE 7: * Marked value is not used for the gamma calculation, but involved for graph The gamma curve shall be approximately linear by converting the measured luminance data to logarithm (axis) The -axis is normalized log 0 (L-L 0 ) where L 0 is the luminance at start point of the exponential gamma curve as 0 IRE in. gamma and the X-axis is just input signal. For example, the logarithmic plot for a gamma curve of Figure 8 is shown in Figure If output luminance and input signals are calculated in logarithmic value between 0 IRE (or 8. IRE for BT.709) and 0 IRE ~ 95 IRE, then the average of the ratios becomes the display gamma (decoding gamma) like Equation (9) Log(Output lumiannce) IRE 0 IRE 0 IRE 30 IRE 40 IRE 50 IRE 60 IRE 70 IRE 80 IRE 90 IRE 00 IRE Input Signal Level Page Doc. 457J SEMI

13 Figure 0 Normalized Logarithmic Luminance Plot In the case of the Constant APL gamma or Constant luminance APL gamma, the gamma is calculated with normalized value by the maximum luminance of the gamma curve. n log Lnorm, k D (9) n k log Inorm, k NOTE 8: n 9 to 8 (0% ~ 5% input step) from 0 IRE to 95 IRE for. gamma or ITU-R BT.709 gamma. NOTE 9: L norm,k Normalized value of the luminance measured at each level, (L measured L 0 ) / (L max L o ) for. gamma (L measured L 8.% ) / (L max L 8.% ) for BT.709 gamma NOTE 0: I norm,k Normalized value of input digital value at each level over 8.% ( of 55) I k /55 for. gamma (I k )/34 for BT Actually, ITU-R BT. 709 gamma equation has offset of and scaling factor of.099, but these are negated by over 8.% input condition and normalization effect respectively Especially, in case of the Constant APL gamma, the gamma values of each APLs might be different depends on APL and should be averaged with each gamma values of 0% ~ 90% APL. NOTE : m generally 9 from 0%, 0%,., 90% APL 6.5 Gamma Accuracy 6.5. Although the gamma is calculated near to., we still don t know the deviation of the gamma values at each level. So, a method to estimate the accuracy of the gamma is necessary. Gamma accuracy (%) for gamma and RGB gamma is defined as follows. NOTE : γ S =., for conventional. gamma NOTE 3: γ S = /0.45, for BT. 709 gamma D accu D, D,... D, m 6.6 Inter-level Gamma 6.6. In general, the gamma is calculated based on the black level (luminance of 0 data input). But, for the purpose of more detail gamma estimation between each level, it needs to get the logarithmic ratio of each neighbored input and output levels in the normalized luminance curve. m S S D. 00 S (0) () D, k log log L I norm, k L norm, k norm, k I norm, k () Page 3 Doc. 457J SEMI

14 NOTE 4: L norm,k Normalized value of the luminance measured at each level, (L measured L 0 ) / (L max L o ) for. gamma (L measured L 8.% ) / (L max L 8.% ) for BT.709 gamma 6.6. Generally, this inter-level gamma values have wider range gamma value comparing with general (black level based) gamma because the black level based gamma has longer input intervals from black to level k. 7 Color Measurement 7. Primary and Secondary Color 7.. This is to measure the luminance and chromaticity coordinates of primary Red, Green, and Blue of 00 IRE at the center of the full screen or in a window box measuring 4% ~ 5% of the full screen. Additionally, it is useful to measure the secondary colors (Cyan, Magenta and ellow) by the same method. 00 IRE Red 00 IRE Green 00 IRE Blue Figure Full Screen Pattern for Primary Color Measurement 7. Primary and Second Color Coordinates 7.. This measures the CIE 93 chromaticity coordinates (x, y) or CIE 976 (u',v') of the primary and secondary colors in color area that a display represents. To do this, a full screen (or partial screen box, depending on APL luminance) of all the primary and secondary colors is displayed and it is done by measuring the luminance and (x, y) or (u',v')chromaticity coordinates at the center of the screen. 7.. From the following additive color mixture equations (Grassman s Law), the standard (x, y) coordinates of these colors at any white balance CCT are available based on ITU-R BT.709 coordinates of D65 white. In the equations, and are gains of mixed 3 colors that are R, G and B respectively where their coordinates are (x, y ), (x, y ) and (x 3, x 3 ). The 3 gains of R, G and B might be 0 to 55 in case of 8 bits integer representation. x y x y y y y y y y 3 x y x y (3a) (3b) Page 4 Doc. 457J SEMI

15 7..3 In table, B93 means 9300K white point in black body locus, and R, G, B are gamma de-corrected values of 8 bits In the same color gamut, after the white balance adjustment of R, G, B by downward gain control, the luminance() is calculated by following equations in the case of same ranges of and R, G, B values like 0~55 or 6~35 : = 0.6R G B for D65 CCT of ITU-R BT.709 (4) = 0.90R G B for 9300K(B93) white balance adjustment of RGB gain (5) Table ITU-R BT. 709 Primary and Secondary Color Coordinate Standards CCT D65 Color x y R G B / D65white (%) 9300K(B93) x y R G B / B93white (%) Red Green Blue Cyan Magenta ellow Page 5 Doc. 457J SEMI

16 BT.709 Gamut Measured Gamut CIE93 Gamut Figure General Pattern of Gamut Measurement 7.3 Primary Color Accuracy and Color Gamut Evaluation 7.3. The measured color coordinates are compared to the ITU-R BT.709 values in order to check the Δu v or Δxy The color coordinates for R, G, B, C, M, need to be in compliance with the ITU-R BT.709 Standard Color Following equation shall be used for calculation of color difference between the measured color coordinate data and reference ITU-R BT. 709 value in CIE 976 UCS coordinate system, as an example of red primary. u' v' R u' u' v' v' R D R D (6) NOTE 5: u D, v D The reference color coordinate of ITU-R BT. 709 primaries Color gamut area is used to evaluate the color range of the display Generally, it is calculated with u v or xy coordinate system. Color gamut related ratios that is relative to the ITU-R.BT.709 standard color gamut are defined as follows. Color Reproduction Ratio (%) Display gamut area 00 Standard gamut area (7) Page 6 Doc. 457J SEMI

17 Overlap area Standard Reproduction Ratio (%) 00 Standard gamut area Standard Reproducti on ratio Color Reproducti on ratio 8 White Balance Measurement 8. The CCT and Δ uv of PDP vary according to APL and input level as shown in the following Figure 3. So, the White Balance of PDP needs to be measured for all APL and all input levels. Further, it is conceptually considered as Gray Balance rather than White Balance. CCT(K) Standard Fidelity (%) Overlap area 00 Display gamut area 0 APL 0 APL 30 APL 40 APL 50 APL 60 APL 70 APL 80 APL 90 APL 00 0 IRE 30 IRE 40 IRE 50 IRE 60 IRE 70 IRE 80 IRE 90 IRE 00 IRE (8) (9a) (9b) Input Signal Δuv IRE 30 IRE 40 IRE 50 IRE 60 IRE 70 IRE 80 IRE 90 IRE 00 IRE Input Signal Figure 3 An Example of CCT and Δuv according to APL and Input Level (Bare PDP Module) 8.. Δuv means the perpendicular deviation from Planckian (black body) locus with plus or minus sign based on a CIE 960 uv chromaticity space, It is different from Euclidean distance Δu v for color difference at CIE 976 u v color space. 8.. From the example of Figure 3, the curves have some fluctuation because they are from a bare PDP module before it is adjusted on a TV board. CCT goes up and Δuv goes down while the level varies high in a APL. But, CCT goes down and Δuv goes up while the APL varies high in a level. 8. Constant APL White Balance Accuracy 8.. White balance is important to display exact colors through all levels because it is the origin of primary colors. So it is effective to describe as concept of accuracy for the standard white points in Table 3. Page 7 Doc. 457J SEMI

18 Table 3 Standards of White Points White CCT Coordinates CIE 93 CIE 976 CCT Δuv (deviation) x y u v D D K(B93) For measuring the white balance accuracy, all level input at all APLs are needed, so the Constant APL gamma pattern is used like Figure or 4. Besides D65, white accuracy (%) is calculated at any white points according to CCT with following equation. m n 000 W accu. 00 u' v' j, k (0) mn j k NOTE 6: m 9 (0, 0,.., 90% APL), 00% APL is meaningless because full screen is 00 IRE. NOTE 7: n 0 (0, 0,.., 00 IRE) or 9 (0, 30,..., 00 IRE) 8.3 White Balance Evaluation at an APL 8.3. For the white balance evaluation of an APL range, Δu v of chromaticity space and ΔE * uv based on relatively perceivable lightness viewpoint are measured ΔE * uv implies the relative luminance deviation added to Δu*v* ΔE * uv is drived from CIELUV(CIE 976 L * u * v * ) color space These are executed like the pattern for the Constant-luminance APL gamma which is made on a center screen box with 4% ~ 5% of total screen area while a box of varying gray level is displayed The input of a center box is varied from 0 IRE to 00 IRE. u' v' WB _ k u' u' v' v' W k W k () NOTE 8: u k, v k the box window coordinate from 0 IRE to 00 IRE * E uv * * * ( L ) ( u ) ( v ) () NOTE 9: ΔL * = L * L *, Δu * = u * u *, Δv * = v * v * NOTE 0: u * * ' 3L ( u W u ' D ) and * * ' v 3L ( v W v ), where u W, v W are for target white and u D, v D are for measured display. ' D We want to know only the white balance deviation not luminance difference, when we assume the maximum luminance up to 4% ~ 5% APL is approximately same to ΔL * = 0. (The gamma is constant up to 4% ~ 5% APL) Also, the 00 IRE white of the window shall be chosen as a relative base coordinate, Similarly, u * = 3L (u W - u W )=0 (3) v * = 0 (4) Then, at the input k of the window, Page 8 Doc. 457J SEMI

19 * NOTE : Lk 6( , where the range of k is from 0 IRE. NOTE : Here, ref is obtained at 00 IRE of the window and k means the window luminance from 0 IRE to 00 IRE. 8.4 Full-screen White Balance k ref ) 3 E ' 6 for * WB _ k k ref ( u 3L * k * k ) ( u ( v ' W * k ) -u ' k 8.4. Generally, full-screen CCT and Δuv have representative values of above all gray points CCT and Δuv including all APLs and input levels. Full-screen data (for varying APL) Figure 4 Pattern for the Full-screen White Balance Measurement ) 0 00% APL ( v ' W -v ' k ) (5) In Figure 4, APL(%) = 00 (/55). (6) = 55 (APL / 00 ) /. (7) 0 CCT APL CCT j 0 (8) j j 0 uv APL uv j 0 (9) NOTE 3: j index for 0, 0,.., 00% APL 9 Color Accuracy Measurement based on a Reference Pattern 9. Geometry and APL of Reference Pattern 9.. Figure 5 is for the color reference pattern from the SMPTE standard. It is presented that the vertical size is 00%, so the horizontal size is 60% in 6:9 pattern which is symmetrical in up-down and left-right. Page 9 Doc. 457J SEMI

20 Figure 5 Color Reference Pattern Geometry (SMPTE 303M-00) 9.. The color area of the pattern is /.78 = 43.7%, so the black background is 56.3% without over-scan The next table is about the color coordinates and 8bit RGB (0 ~ 55) value before gamma correction From the color average the APL of the pattern is as following APL6 :9_ ref % (30) For the 4:3 pattern, it is same as the 6:9 color pattern except the side black panel is 5% instead of 7.5%. The color area is /.33 = 58.5% without over-scan APL4 :3_ ref % (3) Data of Reference Pattern 9.. For getting input data and target color coordinates of the reference pattern, we need to know the tri-stimulate value to RGB conversion matrix for the white color temperature. 9.. The following table is XZ RGB Matrix for D65 & 9300K black body white coordinates based on ITU-R BT.709. About D65, it is referred to SMPTE 303M-00. All color reference patterns have their own spectral reflectance with range from 380nm to 780nm, and are converted to CIE XZ tri-stimulus values for D65 and B93 reference light source. Then, the tri-stimulus values of color reference patterns are finally converted to RGB values for measurement as shown in table 5 and 6. Table 4 XZ RGB Matrices D K (B93) X Z X Z R G B Page 0 Doc. 457J SEMI

21 9..3 The following tables are about color coordinates of the color reference pattern and RGB video data before gamma correction for D65 and B93 based on each XZ RGB matrices. Table 5 Color Reference Pattern based on D65 and ITU-R BT. 709 primaries Color no.(k) Color Patch u v R G B Dark skin Light skin Blue sky Foliage Blue flower Bluish green Orange Purple blue Moderate red Purple ellow green Orange yellow Blue Green Red ellow Magenta Cyan White Neutral Neutral Neutral Neutral Black Ave Table 6 Color Reference Pattern based on 9300K and ITU-R BT. 709 primaries Color no.(k) Color Patch u v R G B Dark skin Light skin Blue sky Foliage Page Doc. 457J SEMI

22 5 Blue flower Bluish green Orange Purple blue Moderate red Purple ellow green Orange yellow Blue Green Red ellow Magenta Cyan White Neutral Neutral Neutral Neutral Black Ave The next table is about the gamma corrected video input values when the pattern generator should be inputted with integer 8 bit R, G and B instead of original color reference pattern scene. From this table, the pre-gamma APL is.56% for. gamma, 6:9 screen, while the post-gamma APL is.88% Mathematically, by Cauchy-Schwarz inequality, the average of exponential γ D of positive values is greater than or equal to the exponential γ D of average of the values when γ D >. The pre-gamma APL with general video level s distribution, it might be roughly times of post-gamma APL. APL n pre APL post D D ( ) ( pixel k / 55) n k NOTE 4: APL pre pre-gamma APL (%) NOTE 5: APL post post-gamma APL (%) NOTE 6: the equal in the case of single data or 0/00 IRE all around full screen NOTE 7: n number of pixel in the screen E NOTE 8: pixel k 8 bit pixel value, ( original k / 55) ) 55 (3) Page Doc. 457J SEMI

23 Table 7 Color Reference Pattern Integer Inputs for D65 ITU-R BT. 709 Primaries Col. no. (k) 8 bit Inputs (. Gamma corrected). Gamma de-corrected R G B R G B Ave Table 8 Color Reference Pattern Outputted by Integer Input of D65 ITU-R BT. 709 Primaries Col. no. (k) D65 White Balance adjusted B93 White Balance adjusted u v / D65white (%) u v / B93white (%) Page 3 Doc. 457J SEMI

24 Col. no. (k) D65 White Balance adjusted B93 White Balance adjusted u v / D65white (%) u v / B93white (%) Ave Figure 6 Test Pattern Management for Maintaining APL 9..6 Table 8 is about the output color coordinates by the gamma corrected D65 integer inputs for D65 and B93 white balance, it is a little different with the original coordinates of the color reference pattern because of the integer gamma corrected inputs of 8 bits In this table, B93 target values are calculated by white balance gain control from D65 RGB inputs instead of B93 XZ RGB table because display input data shout be same irrespective of a display CCT mode For measuring convenience of the same position, screen uniformity and the same APL of the color reference pattern, it can be done as Figure 6. Page 4 Doc. 457J SEMI

25 9..9 For the color reference pattern, the average color accuracy (%) is calculated as follows: C accu_k = x Δu v k (33) CRP_C C _ (34) 9..0 The average ΔE* accuracy (%) is calculated as follows: ΔE * accu_k = 00 - ΔE * k (35) * * CRP_ E E _ (36) 0 Report 0. For example, the measurement results are reported as given in Table. Table 9 Example of tone and color measurement results Display model no. TEST 00-0 Gamma accuracy (%) Primary/ Secondary color & Gamut White balance accuracy R G B White γ D γ accu. γ D γ accu. γ D γ accu. γ D γ accu Color Reference Measurement Color Standard Standard Reproduction Reproduction Fidelity (%) u v u v Ratio (%) Ratio (%) R G B C M Reference D65 A number of APL Accuracy (%) u v Color accuracy (%) CRP_ΔE* accu (D65) 95.4 CRP_ΔE* accu (B93) 94. NOTICE: Semiconductor Equipment and Materials International (SEMI) makes no warranties or representations as to the suitability of the Standards and Safety Guidelines set forth herein for any particular application. The determination of the suitability of the Standard or Safety Guideline is solely the responsibility of the user. Users are cautioned to refer to manufacturer s instructions, product labels, product data sheets, and other relevant literature, respecting any materials or equipment mentioned herein. Standards and Safety Guidelines are subject to change without notice. By publication of this Standard or Safety Guideline, SEMI takes no position respecting the validity of any patent rights or copyrights asserted in connection with any items mentioned in this Standard or Safety Guideline. Users of this Standard or Safety Guideline are expressly advised that determination of any such patent rights or copyrights, and the risk of infringement of such rights are entirely their own responsibility. Page 5 Doc. 457J SEMI