SMPTE RP 2077-2013 Full-Range Image Mapping.pdf

《SMPTE RP 2077-2013 Full-Range Image Mapping.pdf》由会员分享，可在线阅读，更多相关《SMPTE RP 2077-2013 Full-Range Image Mapping.pdf（18页珍藏版）》请在麦多课文档分享上搜索。

1、Page 1 of 18 Approved November 7, 2013 Copyright 2013 by THE SOCIETY OF MOTION PICTURE AND TELEVISION ENGINEERS 3 Barker Avenue, White Plains, NY 10601 (914) 761-1100 SMPTE RP 2077:2013 SMPTE RECOMMENDED PRACTICE Full-Range Image Mapping Page 1 of 18 pages Table of Contents Foreword 2 Intellectual P

2、roperty . 2 Introduction 2 1 Scope 5 2 Conformance Notation 5 3 Normative References 5 4 Terms and Definitions . 6 5 Mapping to Interfaces and Formats that Rely upon Protected Code Values . 6 6 Mapping to Interfaces and Formats with Narrow-Range Image . 7 Annex A Bibliography (Informative) 8 Annex B

3、 Mapping to and from a 10-Bit Interface that Relies upon Protected Code Values (Informative) . 9 Annex C Example Mappings to and from a Format with Narrow-Range Image (Informative) 10 SMPTE RP 2077:2013 Page 2 of 18 pages Foreword SMPTE (the Society of Motion Picture and Television Engineers) is an

4、internationally-recognized standards developing organization. Headquartered and incorporated in the United States of America, SMPTE has members in over 80 countries on six continents. SMPTEs Engineering Documents, including Standards, Recommended Practices, and Engineering Guidelines, are prepared b

5、y SMPTEs Technology Committees. Participation in these Committees is open to all with a bona fide interest in their work. SMPTE cooperates closely with other standards-developing organizations, including ISO, IEC and ITU. SMPTE Engineering Documents are drafted in accordance with the rules given in

6、Part XIII of its Operations Manual. SMPTE RP 2077 was prepared by Technology Committee 10E. Intellectual Property At the time of publication no notice had been received by SMPTE claiming patent rights essential to the implementation of this Engineering Document. However, attention is drawn to the po

7、ssibility that some of the elements of this document may be the subject of patent rights. SMPTE shall not be held responsible for identifying any or all such patent rights. Introduction This section is entirely informative and does not form an integral part of this Engineering Document. Each pixel i

8、n a digital image is commonly encoded as a triplet of n-bit binary words see, for instance, SMPTE ST 125, SMPTE ST 274 and ST SMPTE 296 for the process by which such encoding is generated. In a full-range digital image, the code values of the reference black and white triplets are the minimum and ma

9、ximum values that can be encoded by an n-bit unsigned binary integer, (0, 0, 0) and (2n 1, 2n 1, 2n 1) respectively, i.e. the “full range“ of code values. This specification defines mappings allowing source full-range images to be carried by (i) digital interfaces that rely upon protected values, e.

10、g. for synchronization, and (ii) narrow-range digital image formats where the full-range of code values is not utilized for the range between reference black and white11The terms “reference black“ and “reference white“ have the same meanings herein as in SMPTE ST 125, SMPTE ST 274 and SMPTE ST 296,

11、which state “picture information shall be linearly represented by red, green and blue tristimulus values (RGB), lying in the range 0 (reference black) to 1 (reference white)“. . These mappings can be used to devise methods to interoperate with analog workflows, but those methods are outside the scop

12、e of this document. On digital interfaces that rely upon protected values, a small number of code values are prohibited from appearing in triplets. For instance, the 10-bit interface defined in SMPTE ST 292-1 protects code values outside the range 4,1019. SMPTE RP 2077:2013 Page 3 of 18 pages 410194

13、10191023010230Full-Range ImageInterfacewith Protected Code ValuesFull-Range ImageFigure 1 Full-Range Image Mapping over an Interface that Rely upon Protected Code Values In this first scenario, Section 4.1 specifies that, given a full-range image as input, code values falling within the protected ra

14、nge are clipped to the closest permitted value without any scaling. As illustrated in Figure 1, no additional processing is required if the resulting narrow-range image is mapped back to full-range at a later time, but the resulting full-range image will be missing any code value clipped by the init

15、ial mapping to the digital interface. In narrow-range digital image formats, the reference black and white triplets, (RBLACK, RBLACK, RBLACK) and (RWHITE, RWHITE, RWHITE) respectively, are specified with RBLACK0 and RWHITEb The function Mpermittedfull(“map from permitted range to full-range“) shall

16、map a code value CVPon the interface or format to a code value CVFULLfor the full-range image according to Mpermittedfull: CVP CVFULL= CVP (2) Note: Any information in a full-range image encoded using code values outside the range CVLOW, CVHIGH is lost in this mapping. Equivalently, full-range image

17、s sourced from an interface or format with protected code values will not contain code values outside the range CVLOW, CVHIGH. Annex B contains an illustrative mapping to and from a 10-bit interface with CVLOW = 4and CVHIGH= 1019. SMPTE RP 2077:2013 Page 7 of 18 pages 6 Mapping to Interfaces and For

18、mats with Narrow-Range Image The following applies when mapping a full-range image to a format or interface that uses narrow-range images of the same bit length n. The function Mfull narrow(“map from full-range to narrow-range“) shall map a code value CVFULLfrom a full-range image to a narrow-range

19、code value CVNon the interface or format according to: Mfull narrow: CVFULL CVN= floor ( 0.5 + CVFULL/ (2n-1) (RWHITE RBLACK) + RBLACK), (3) where floor(x) is equal to the largest integer not greater than x. The function Mnarrowfull(“map from narrow-range to full-range“) shall map a code value CVNfr

20、om the interface or format to a full-range code value CVFULLaccording to Mnarrowfull: CVN CVFULL= floor ( 0.5 + clamp ( 0.0, 1.0, (CVN- RBLACK) / (RWHITE RBLACK) ) (2n-1) ) , (4) where floor(x) and clamp(a, b, x) use the same definitions as in the previous section. Notes: 1 floor(0.5 + x) expresses

21、rounding of positive x to the nearest integer. 2 No slopes, curves, knee or shoulder are specified at the top and bottom of the mapping interval. 3 The pair of equations above are constructed such that Mfull narrow(Mnarrow full( Mfull narrow( x ) ) ) = Mfullnarrow( x ), i.e. once a full-range value

22、has been mapped to a narrow-range value (and quantization information necessarily lost), no further information is lost if the mapping is repeated. 4 CVNin Equation (4) can be smaller than RBLACKor larger than RWHITE, but any information outside the range RBLACK, RWHITE is lost when mapping from nar

23、row-range to full-range using Mnarrowfull, i.e. Mnarrowfullclips code-values outside the range RBLACK, RWHITE. If such clipping is not desired, code values outside the range RBLACK, RWHITE can be mapped to the range RBLACK, RWHITE prior to applying CNF, using a mapping outside the scope of this spec

24、ification. 5 The Mfullnarrowand Mfullpermittedmappings are not commutative. If an interface or format uses both protected values and narrow-range image, then the Mfullnarrowmapping is applied first since the Mfullpermittedmapping applies only to interfaces and formats with white and black reference

25、code values equal to that of full-range. Conversely, the inverse functions are applied in the opposite order, i.e., the Mpermittedfullmapping is applied first, followed by Mnarrowfull. Annex C contains example mappings according to Equations (3) and (4), for both 8-bit (RBLACK= 16 and RWHITE= 235) a

26、nd 10-bit images (RBLACK= 64 and RWHITE= 940). SMPTE RP 2077:2013 Page 8 of 18 pages Annex A Bibliography (Informative) Note: All references in this document to other SMPTE documents use the current numbering style (e.g. SMPTE ST 274:2008) although, during a transitional phase, the document as publi

27、shed (printed or PDF) may bear an older designation (such as SMPTE 274M-2008). Documents with the same root number (e.g. 274) and publication year (e.g. 2008) are functionally identical. SMPTE ST 125:2013, SDTV Component Video Signal Coding 4:4:4 and 4:2:2 for 13.5 MHz and 18 MHz Systems SMPTE ST 27

28、4:2008, Television 1920 x 1080 Image Sample Structure, Digital Representation and Digital Timing Reference Sequences for Multiple Picture Rates SMPTE ST 292-1:2012, 1.5 Gb/s Signal/Data Serial Interface SMPTE ST 296:2012, 1280 x 720 Progressive Image 4:2:2 and 4:4:4 Sample Structure Analog and Digit

29、al Representation and Analog Interface SMPTE RP 291-2:2013, Ancillary Data Space Use 4:2:2 SDTV and HDTV Component Systems and 4:2:2 2048 x 1080 Production Image Formats SMPTE RP 2077:2013 Page 9 of 18 pages Annex B Mapping to and from a 10-Bit Interface that Relies upon Protected Code Values (Infor

30、mative) The following illustrates the mapping of a 10-bit full-range image to and from a 10-bit interface that relies upon protected code values 0, 3 1020,1023, using the Mfullpermittedand Mpermittedfullmappings, respectively. The same protected code values are, for instance, used by the digital int

31、erface specified in SMPTE ST 292-1. Table B.1 Mapping code values from a full-range image (CVFULL) to a 10-bit interface (CVP) that rely upon protected code values CVFULLCVP0 4 1 4 2 4 3 4 4 4 5 5 127 127 128 128 1018 1018 1019 1019 1020 1019 1021 1019 1022 1019 1023 1019 Table B.2 Mapping from a 10

32、-bit interface (CVP) that rely upon protected code values to a full-range image (CVFULL) CVPCVFULL4 4 5 5 6 6 127 127 128 128 1017 1017 1018 1018 1019 1019 SMPTE RP 2077:2013 Page 10 of 18 pages Annex C Example Mappings to and from a Format with Narrow-Range Image (Informative) The following tables

33、illustrate the mapping of code values for a full-range image to and from code values for narrow-range image, using the Mfull narrowand Mnarrowfullmappings, respectively. The black and white reference triplets used in these examples are those used, for instance, in the RGB digital representation spec

34、ified in SMPTE ST 274. As indicated in Section 6, Table C.2 and Table C.4 perform a mapping from narrow-range to full-range in which code values below reference black and above reference white are lost. If conservation of these values is required, another mapping can be used. Table C.1 Mapping code

35、values from a 8-bit full-range image (CVFULL) to an 8-bit image (CVN) with black reference triplet (16, 16, 16) and white reference triplet (235, 235, 235) CVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVN0 16 32 43 64 71 96 98 128 126 160 153 192 181 224 208 1 17 33 44 65 72

36、97 99 129 127 161 154 193 182 225 209 2 18 34 45 66 73 98 100 130 128 162 155 194 183 226 210 3 19 35 46 67 74 99 101 131 129 163 156 195 183 227 211 4 19 36 47 68 74 100 102 132 129 164 157 196 184 228 212 5 20 37 48 69 75 101 103 133 130 165 158 197 185 229 213 6 21 38 49 70 76 102 104 134 131 166

37、 159 198 186 230 214 7 22 39 49 71 77 103 104 135 132 167 159 199 187 231 214 8 23 40 50 72 78 104 105 136 133 168 160 200 188 232 215 9 24 41 51 73 79 105 106 137 134 169 161 201 189 233 216 10 25 42 52 74 80 106 107 138 135 170 162 202 189 234 217 11 25 43 53 75 80 107 108 139 135 171 163 203 190

38、235 218 12 26 44 54 76 81 108 109 140 136 172 164 204 191 236 219 13 27 45 55 77 82 109 110 141 137 173 165 205 192 237 220 14 28 46 56 78 83 110 110 142 138 174 165 206 193 238 220 15 29 47 56 79 84 111 111 143 139 175 166 207 194 239 221 16 30 48 57 80 85 112 112 144 140 176 167 208 195 240 222 17

39、 31 49 58 81 86 113 113 145 141 177 168 209 195 241 223 18 31 50 59 82 86 114 114 146 141 178 169 210 196 242 224 19 32 51 60 83 87 115 115 147 142 179 170 211 197 243 225 20 33 52 61 84 88 116 116 148 143 180 171 212 198 244 226 21 34 53 62 85 89 117 116 149 144 181 171 213 199 245 226 22 35 54 62

40、86 90 118 117 150 145 182 172 214 200 246 227 23 36 55 63 87 91 119 118 151 146 183 173 215 201 247 228 24 37 56 64 88 92 120 119 152 147 184 174 216 202 248 229 25 37 57 65 89 92 121 120 153 147 185 175 217 202 249 230 26 38 58 66 90 93 122 121 154 148 186 176 218 203 250 231 27 39 59 67 91 94 123

41、122 155 149 187 177 219 204 251 232 28 40 60 68 92 95 124 122 156 150 188 177 220 205 252 232 29 41 61 68 93 96 125 123 157 151 189 178 221 206 253 233 30 42 62 69 94 97 126 124 158 152 190 179 222 207 254 234 31 43 63 70 95 98 127 125 159 153 191 180 223 208 255 235 SMPTE RP 2077:2013 Page 11 of 18

42、 pages Table C.2 Mapping code values from an 8-bit image (CVN) with black reference triplet (16, 16, 16) and white reference triplet (235, 235, 235) to an 8-bit full-range image (CVFULL) CVN CVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULL0 0 32 19 64 56 96 93 128 130 160 168 19

43、2 205 224 242 1 0 33 20 65 57 97 94 129 132 161 169 193 206 225 243 2 0 34 21 66 58 98 95 130 133 162 170 194 207 226 245 3 0 35 22 67 59 99 97 131 134 163 171 195 208 227 246 4 0 36 23 68 61 100 98 132 135 164 172 196 210 228 247 5 0 37 24 69 62 101 99 133 136 165 173 197 211 229 248 6 0 38 26 70 6

44、3 102 100 134 137 166 175 198 212 230 249 7 0 39 27 71 64 103 101 135 139 167 176 199 213 231 250 8 0 40 28 72 65 104 102 136 140 168 177 200 214 232 252 9 0 41 29 73 66 105 104 137 141 169 178 201 215 233 253 10 0 42 30 74 68 106 105 138 142 170 179 202 217 234 254 11 0 43 31 75 69 107 106 139 143

45、171 180 203 218 235 255 12 0 44 33 76 70 108 107 140 144 172 182 204 219 236 255 13 0 45 34 77 71 109 108 141 146 173 183 205 220 237 255 14 0 46 35 78 72 110 109 142 147 174 184 206 221 238 255 15 0 47 36 79 73 111 111 143 148 175 185 207 222 239 255 16 0 48 37 80 75 112 112 144 149 176 186 208 224

46、 240 255 17 1 49 38 81 76 113 113 145 150 177 187 209 225 241 255 18 2 50 40 82 77 114 114 146 151 178 189 210 226 242 255 19 3 51 41 83 78 115 115 147 153 179 190 211 227 243 255 20 5 52 42 84 79 116 116 148 154 180 191 212 228 244 255 21 6 53 43 85 80 117 118 149 155 181 192 213 229 245 255 22 7 5

47、4 44 86 82 118 119 150 156 182 193 214 231 246 255 23 8 55 45 87 83 119 120 151 157 183 194 215 232 247 255 24 9 56 47 88 84 120 121 152 158 184 196 216 233 248 255 25 10 57 48 89 85 121 122 153 160 185 197 217 234 249 255 26 12 58 49 90 86 122 123 154 161 186 198 218 235 250 255 27 13 59 50 91 87 1

48、23 125 155 162 187 199 219 236 251 255 28 14 60 51 92 88 124 126 156 163 188 200 220 238 252 255 29 15 61 52 93 90 125 127 157 164 189 201 221 239 253 255 30 16 62 54 94 91 126 128 158 165 190 203 222 240 254 255 31 17 63 55 95 92 127 129 159 167 191 204 223 241 255 255 SMPTE RP 2077:2013 Page 12 of

49、 18 pages Table C.3 Mapping code values from a 10-bit full-range image (CVFULL) to a 10-bit image (CVN) with black reference triplet (64, 64, 64) and white reference triplet (940, 940, 940). CVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVNCVFULLCVN0 64 128 174 256 283 384 393 512 502 640 612 768 722 896 831 1 65 129 174 257 284 385 394 513 503 641 613 769 722 897 832 2 66 130 175 258 285 386 395 514 504 642 614 770 723 898 833 3 67 131 176 259 286 387 395 5