SMPTE RDD 22-2012 Film Transfer 2048x1556 Image Container and Signal Interface.pdf

1、Copyright 2012 by THE SOCIETY OFMOTION PICTURE AND TELEVISION ENGINEERS3 Barker Avenue, White Plains, NY 10601(914) 761-1100Approved March 12, 2012Film Transfer 2048x1556 Image Containerand Signal InterfacePage 1 of 56 pagesThe attached document is a Proposed Registered Disclosure Document prepared

2、by the proponent identifiedbelow. It has been examined by the appropriate SMPTE Technology Committee and is believed to containadequate information to satisfy the objectives defined in the Scope, and to be technically consistent.This document is NOT a Standard, Recommended Practice or Engineering Gu

3、ideline, and does NOT imply afinding or representation of the Society.Errors in this document should be reported to the proponent identified below, with a copy to engsmpte.org.All other inquiries in respect of this document, including inquiries as to intellectual property requirements thatmay be att

4、ached to use of the disclosed technology, should be addressed to the proponent identified below.Proponent contact information:Tadashi Okano,Sony Corporation,4-14-1 Asahi-cho, AtsugiKanagawa, 243-0014JapanEmail: SMPTE RDD 22:2012SMPTE REGISTEREDDISCLOSURE DOCUMENTSMPTE RDD 22:2012Page 2 of 56 pagesTa

5、ble of Contents PageIntroduction 41 General. 42 Source Image Format . 42.1 F-S Colorimetry 52.1.1 Encoding Primaries Free Scale Gamut (FS-Gamut) 52.1.2 FS-Log curve 62.1.3 Notation of parameter values 82.1.4 Color Ancillary Data 92.2 RGB/Y,CR,CBColorimetry 92.3 Image Pixel Array Representation. 92.3

6、.1 FS-Log Pixel Array Representation . 92.3.2 RGB Pixel Array Representation. 93 Transport Image Container Parameters. 94 Virtual Interface data values and Structure 105 Serial Interface Characteristics 105.1 Timing Reference Signal Data Structure . 135.2 Channel Coding 145.3 Coaxial Cable Interface

7、 155.3.1 Signal Levels and Specifications. 155.4 Connector and Cable Types . 176 Dual Link Image Data Mapping . 177 3 Gb/s Single-Link Mapping 208 Audio. . 208.1 Structure of Audio Data Packet. 248.2 Structure of User Data Words (UDW) . 258.2.1 CLK (audio clock phase data) . 258.2.2 ECC (Error corre

8、ction codes) 298.3 Multiplexing of Audio Data Packet. 308.3.1 Audio Control Packet 308.3.2 Structure of User Data Words (UDW) 318.3.3 Multiplexing of the Audio Control Packets . 359 Time Code 359.1 25 Frame Operation . 359.2 24, 24/1.001 Frame Operation 359.3 Linear Time Code (LTC) Structure 359.3.1

9、 Codeword format 359.3.2 Codeword data content. 359.3.3 Flag bits 359.3.4 Timing of the codeword relative to the image 35SMPTE RDD 22:2012Page 3 of 56 pages9.4 Vertical Interval Timecode (VITC)369.4.1 Codeword format.369.4.2 Codeword data content .369.4.3 Time address 369.4.4 Flag bits 369.4.5 Binar

10、y groups 369.4.6 Interface-Field flag.369.5 Format of Ancillary Time Code Packets .389.6 Ancillary Time Code Packet Layout .389.7 Format of User Data Words in Ancillary Time Code Packet399.7.1 Bit Assignments within the UDW.399.8 Distributed Binary Bits (DBB).399.8.1 DBB1 Payload Type .399.8.2 Mappi

11、ng of the Time Code Data into Ancillary Data Packets409.8.3 Transmission of ancillary time code packets 429.8.4 ATC Packet Transmission Rate.439.8.5 Ancillary Time Code Packet Location.4310 Payload Identifier.4310.1 Ancillary Data Specification .4311 Color ANC Data Packet .4811.1 Structure of Color

12、ANC4811.2 Color ANC Packet UDW 4811.2.1 ACT1 Color Primary and Reference White Parameter4811.2.2 ACT2 Nonlinear Equation Specifier4911.3 RFS, GFS, BFS, Reference White Primaries, , , , , kexp and LB1, LB1, LB2, LB2, LB3, LB3, LC1, LC1, LC2, LC2, LC3 LC3 parameters 51SMPTE RDD 22:2012Page 4 of 56 pag

13、esIntroductionThe intended application of this RDD is for the transfer of full frame film images to electronic media.This RDD defines an image pixel array and interface container to transport a maximum pixel array of2048 x 1556, over a real time serial digital interface of 1650 lines. The source ima

14、ge is a 4:4:4 image,transported by a dual link interface.The 2048 x 1556 source image is mapped into 2 serial Interfacecontainers, 2 x 1536 x 778 pixels (dual channel) per link. The Y/G image samples are interleaved between theY and C channels of link A. The Red image samples are interleaved between

15、 the Y and C channels of link B.The Blue samples are interleaved between the C channels of link A and the Y channel of link BThis RDD serial interface data structure, sampling frequencies, and synchronizing signals have a lot incommon to other SMPTE defined HDTV Serial interfaces. These commonalitie

16、s should allow for the use ofHDTV distribution equipment currently in use to transport these images. Because of the mapping format,processing and storage of the serial interface data will be required to display the 2048 x 1556 images.1 GeneralThroughout this document reference is made to Ancillary D

17、ata Packets (ANC).These data packets areformatted in conformance with SMPTE ST 291.In the Audio section of the document reference is made to AES-3. Readers may obtain a copy of AES-3 fromthe Audio Engineering Society.The time code defined in this RDD is based on SMPTE ST 12M-1, SMPTE ST 12M-2 and Re

18、commendationITU-R BR.780-2. Details concerning typical physical interfaces, time code data structure, and otherimplementation issues may be found in these references. This RDD does not use all the user defined binarygroup data contained within the time code structure.2 Source Image FormatThe source

19、image format is a progressive capture with a maximum 2048 x 1556 pixel array.The frame rate of the source image array shall be 24, 24/1.001, or 25 frames per second.Table 1 System NumbersSystemNumberFrame Rate Frames per second1 242 24/1.0013 25The pixel array defined above shall be placed within an

20、 image container, however the image may not fill thecontainer.SMPTE RDD 22:2012Page 5 of 56 pagesFigure 1 Image ContainerWhen the source image contains non active pixels, these non-active pixels shall be set to 40h.The bit depth of each pixel shall be 10 bits.Image pixel digital values shall not exc

21、eed the following values:Signal level of100 %3ACh(940(10)Signal level of0 %040h(64(10)Head room forOver shoot3ADh-3FBh(941(10)-1019(10)Foot room forUnder shoot004h-03Fh(4(10)-63(10)2.1 F-S Colorimetry2.1.1 Encoding Primaries Free Scale Gamut (FS-Gamut)The FS colorimetry shall use the CIE XYZ tristim

22、ulus values defined in CIE S 014-2/E: 2006 ISO 11664-2:2007 ranging from 2.00000 to +2.00000. Default values of the primaries and reference white shall be asImage ContainerImageContainercentre,ActiveimagecentreHorizontalPixels,2048Vertical Pixels1556Pixel number0.0Active Image Pixels - May fill the

23、containerNon Active Image Pixels - Set to 040hNon Active Image Pixels - Set to 040hCenter ofthe containerANDcenter ofthe imageLine 0Line 1555SMPTE RDD 22:2012Page 6 of 56 pagesdefined in Table 2 and is defined as FS-Gamut. FS-Gamut and FS-Log are identifying names of thedefined color space and the L

24、og curve in this RDD.Table 2 Default values of FS-Gamut2.1.2 FS-Log curveFrom the RFS, GFS, BFS or user defined Ruser, Guser, Buser tristimulus values, three nonlinear primarycomponents RFS, GFS, BFS or R, G, B shall be calculated according to the following FS-Log curve (Equation1) and if necessary

25、the upper and lower three coordinates defined in Table 3. Where nonlinear primarycomponents L shall be 0 L 1. Table 3 Definition of FS-Log curveWhere, L shall be a linear tristimulus value multiplied by a kexp factor as shown in Equation 2.“a linear value” shall mean that each value be linear relati

26、ve to the amount of light. For example, GFS =0.18000 denotes an 18% Gray and GFS = 1.00000 denotes light coming from a 100% reflector. The kexp valueshall denote the overexposure and the underexposure, i.e., the kexp value of greater than 1 indicates theoverexposure and the kexp value of between 0 a

27、nd 1 indicates the underexposure. The default value of thekexp shall be 1.00000.The FS-Log curve shall use the Equation 1 with L from LB1 to LC1 and if necessary in combination withtwo nonlinear curves connecting the upper three coordinates of (LC1, LC1), (LC2, LC2), (LC3, LC3) with L fromLC1 to LC3

28、 and lower three coordinates of (LB1, LB1), (LB2, LB2), (LB3, LB3) with L from LB3 to LB1 as shown inFigure 3. LB1, LB2, LB3 (LB1 LB2 LB3) and LC1, LC2, LC3 (LC1 1/10 the clock rate Upper band edgeA1 2 UI(3Gb/s)/1UI(1.5Gb/s) Timing jitter (Note 1)A2 0.2 UI(3Gb/s)/0.2 UI(1,5Gb/s) Intrinsic jitter (UI

29、 = unit interval)Test signal Color bar test signal (Note 2)n 10 (preferred) Serial clock divided (Note 3) Notes:1 Color bars are chosen as a nonstressing test signal for jitter measurements. Use of a stressing signal with longruns of zeros may give misleading results.2 Use of a serial clock divider

30、value of 10 may mask word correlated jitter components.3 See SMPTE RP 184 for definition of terms.SMPTE RDD 22:2012Page 17 of 56 pages5.4 Connector and Cable TypesThe male and female connectors shall be 75-ohm BNC as defined in IEC 61169-8, Part 8, Annex A.Application of this RDD does not require a

31、particular type of coaxial cable. It is necessary for the frequencyresponse of the coaxial cable loss, in decibels, to be approximately proportional to 1/f from 1 MHz to the clock frequency of the signal being transmitted to ensure correct operation of automatic cable equalizers overmoderate to maxi

32、mum lengths.Return loss of the correctly terminated transmission line shall be greater than 15 dB over a frequency range of5 MHz to the clock frequency of the signal being transmitted.6 Dual Link Image Data MappingThe source image format is a 2048 x 1556 pixel array this pixel array shall be mapped

33、into two 1536 x 778interface containers per each link. See Figure 8.SMPTE RDD 22:2012Page 18 of 56 pagessamplingrate=74.25MHzG2040G2041G2042G2044G2045G2046EAV (3FFh)EAV (000h)EAV (000h)EAV (XYZh)LN0LN1CRC0CRC1SAV (3FFh)SAV (000h)SAV (000h)SAV (XYZh)G0G1G2G4B2040B2041G2043B2044B2045G2047EAV (3FFh)EAV

34、 (000h)EAV (000h)EAV (XYZh)LN0LN1CRC0CRC1SAV (3FFh)SAV (000h)SAV (000h)SAV (XYZh)B0B1G3B4R2040B2042B2043R2044B2046B2047EAV (3FFh)EAV (000h)EAV (000h)EAV (XYZh)LN0LN1CRC0CRC1SAV (3FFh)SAV (000h)SAV (000h)SAV (XYZh)R0B2B3R4R2041R2042R2043R2045R2046R2047EAV (3FFh)EAV (000h)EAV (000h)EAV (XYZh)LN0LN1CRC

35、0CRC1SAV (3FFh)SAV (000h)SAV (000h)SAV (XYZh)R1R2R3R523.98P sFWORD#23.98P sFWORD#18701871187218731874187512345678933533633733833934034134234325PsFWORD#25PsFWORD#1795179617971798179918001234567892602612622632642652662672671.5GLinkA 1.5GLinkBAudio=16ch:31word/Groupx4Groupx2sample(max)/line=248Word(max

36、)/Line48kHz1-16chMUX/96kHz1-8chMUXYChnCChnANC/AudioData=327Words(23.98PsF)=252Words(25PsF)YChnCChnReserved48kHZBlank/96kHzAudio9-16chFigure13DualLinkImageSampleMappingoftheYandCChannelsSMPTE RDD 22:2012Page 19 of 56 pagesFigure 14 Dual Link Mapping RuleFigure 15 Dual Link Data StreamThe interface sa

37、mpling frequency shall be 74.25 MHz or 74.25/1.001 MHz. for Link A and Link B.The image source line numbers1-778 shall be mapped to interface line numbers 16 thru 793inclusive (Field1). Image source line numbers779 thru 1556 shall be mapped to interface line numbers 841-1618 inclusive(Field 2). See

38、Figure 8.The Y channel of link A shall contain all G image samples except where the sample number modulo 4 =3. Forexample, 0,1,2,4,5,6,8.2046. G Samples 3,7,11,.2047shall be carried in the C Channel. See Figures 13and 14.The C channel of link A shall carry the B image samples as sample pairs 0-1, 4-

39、5, 8-9.2046-2047. SeeFigures 13 and 14.The Y channel of Link B shall carry single R samples interleaved with B sample pairs. R samples,0,4,8,12.2040,2044. The B image sample pairs 2-3,6-7,10-11.2042-2043, 2046-2047. See Figures 13 and14.The C channel of link B shall carry R samples 1,2,3,5,6,7,92045

40、, 2046, 2047. See Figures 13 and 14.The PsF word number as indicated in Figure 13 is the word count for one complete horizontal line. Figure 6represents the sample count of a horizontal line.SMPTE RDD 22:2012Page 20 of 56 pages7 3 Gb/s Single-Link MappingThe interface sampling Frequency shall be 148

41、.5 MHz or 148.5/1.001 MHz.Figure 16 3 Gb/s Serial Data StreamTwo parallel 10-bit interfaces of the same line and frame structure having bit synchronization shall be mappedinto a 20-bit Virtual Interface, consisting of two data streams data stream one and data stream two. Theimage samples from data s

42、tream one and data stream two shall alternate, with the samples from data stream2 appearing first. See Figure 16.8 AudioThis RDD defines the mapping of 24-bit AES digital audio data and associated control information into thehorizontal ancillary data space of the serial digital video interface defin

43、ed in this RDD. The audio data arederived from AES3, hereafter referred to as AES audio. This RDD constrains the content carried by the AES3interface to linear coded samples.Audio sampled at a clock frequency of either 48 kHz or 96 kHz locked (synchronous) to video.Audio channels are transmitted in

44、groups of four, up to a maximum of 16 audio channels in the case of 48-kHzsampling, and up to a maximum of 8 audio channels in case of 96-kHz sampling. Each group is identified by aunique ancillary data ID.Audio data packets are multiplexed (embedded) into the horizontal ancillary data space of the

45、C data stream,and audio control packets are multiplexed into the horizontal ancillary data space of the Y data stream.The modes of transmission carried in an audio data packet shall be the TWO-CHANNEL MODE at a samplingfrequency of 48 kHz, and the SINGLE CHANNEL DOUBLE SAMPLING FREQUENCY MODE at the

46、 samplingSMPTE RDD 22:2012Page 21 of 56 pagesfrequency of 96 kHz as defined in AES3. Audio data channels 14 (CH1 CH4) carry two AES audio channelpairs (AES1 channel 1 and 2 and AES2 channel 1 and 2) in the case of 48 kHz sampling. For 96 kHzsampling, two successive samples of two AES audio channels

47、(AES1 channel 1 1st and 2nd sample andAES2 channel 1 1st and 2nd sample) shall be carried.The 48-kHz sampling audio data derived from two channel pairs shall be configured in an audio data packetas shown in Figure 17. Both channels of a channel pair are derived from the same AES audio source. Thenum

48、ber of samples per channel used for one audio data packet shall be constant and is equal to one. Thenumber of audio data packets in a given group shall be less than or equal to Na in a horizontal ancillary datablock.Figure 18 shows the audio data packet at the sampling rate of 96 kHz. AES sub frames

49、 1 and 2 carrysuccessive samples of the same AES audio signal. Both channels shall be derived from the same AES audiosource. The number of samples per channel used for one audio data packet shall be constant and equal totwo. The number of audio data packets in a given group is less than or equal to Na/2 in a horizontal ancillarydata block.Two types of ancillary data packets carrying AES audio information are defined. Each audio data packet shallcarry all of