Overview of Digital Video Standards Naohiro Shichijo Interfaculty Initiative in Information Studies The University of Tokyo <shichi@iii.u-tokyo.ac.jp> This session is titled “Overview of Digital Video Standards”

Digital Video Standards Focuses on professional digital video standards (ie. non-consumer) Describe following items briefly: Video Signal Sampling Encoding Transferring Tape recording In this 90 minutes tutorial. I’d like to introduce the concept and implementation of various digital video systems. We focus on especially “professional digital video systems.” That is different from ordinary consumer video systems. But they resembles in many points. Because consumer video systems are descendent of professional video systems. So, the understanding of professional video systems will likely to help to understand consumer video systems. In order to understand digital video systems, we have to know about these concepts, That is. Video signal etc… 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Roadmap Analog Video Digital Video HDTV Digital Video Recording So, This tutorial is composed of 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Digital Video at Work -1 Contents creation in analog/digital television. Cinema production Computer graphics Beforehand I’d like to show you a few applications of digital video systems. Of course, the digital video systems are used to create contents of television. The terrestrial TV broadcasting is done in analog in many places of the world. But the contents creation is done mostly in digital. Digital video systems introduced small camera for ENG (Electric News Gathering). Especially for NEWS program, the digital systems are widely spread. And also, for Cinema. In these days, the most of the commercial films are edited in digital. As you may know, the Starwars Episode II is shooted in HDTV camera, and edited using HD-D5 systems. And in some places, they are projected using DLP cinema projectors. The third example is computer graphics. In computer graphics creation, uncompressed component systems like D-1 or D-6 is used. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Digital Video at Work -2 DTV Advanced Television System (USA, Taiwan) http://www.atsc.org/ DVB Digital Video Broadcasting (Europe) http://www.dvb.org/ ISDB Integrated Services Digital Broadcating (Japan) http://www.arib.or.jp/ In these days, The terrestrial broadcasting and satellite broadcasting are changing from analog to digital. These are some examples of the efforts to introduce digital television system in the world. The first one is Advanced Television in united states. DVB is used mainly in Europe. The ISDB is the standard in Japan. 11/14/2018 Naohiro Shichijo, The University of Tokyo

History of Digital Video -0 Around 1970: Digital Black-box Systems Frame synchronizer Standards converter Timebase corrector  Analog I/O Around 1980: Digital VTR Systems DVE, Graphic system, Still store  very limited interoperability 11/14/2018 Naohiro Shichijo, The University of Tokyo

History of Digital Video Standards (1/2) 1982: ITU-R Rec. BT. 601 Encoding parameters for digital television for studios (describes 4:2:2 encoding) 1986: ITU-R Rec. BT. 657 Digital television tape recording. Standards for the international exchange of television programs on magnetic tape  SMPTE 224M~228M (known as D-1) 1990: SMPTE 224M~248M Composite Digital Encoding for NTSC(known as D-2) 1993: IEC61179 19mm Type D-2 Composite Format (includes NTSC, PAL) 1993: SMPTE 264M,265M ½ -in Type D-3 Composite Format 525/60, 625/50 1996: SMPTE 279M ½ -in Type D-5 Component Format – 525/60 and 625/50 11/14/2018 Naohiro Shichijo, The University of Tokyo

History of Digital Video Standards (2/2) 1995: SMPTE 227M,228M Television Digital Recording 19mm – Type D-6 1998: SMPTE 306M,307M,314M Television Digital Recording 6.35mm – Type D-7 Component Format 25Mbit/s, 50Mbit/s (also known as DVCPRO) 1999: SMPTE 316M,317M Television Digital Recoding 12.65mm – Type D-9 (also known as Digital-S) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Introduction to Analog Video

Naohiro Shichijo, The University of Tokyo TV Standards NTSC PAL SECAM 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo 11/14/2018 Naohiro Shichijo, The University of Tokyo

Standard Definition of Composite Video Parameters 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Colorimetry SMPTE CIE x-y color chart By cietoppm.c 11/14/2018 Naohiro Shichijo, The University of Tokyo

Luminance and Chromaticity Tristimulus chromaticity coordinate system, (X, Y, Z) x=X/(X+Y+Z), y=Y/(X+Y+Z) Color is represented in the CIE2002 primarily colors. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Basic ingredients Brightness information Chrominance information (scaled to prevent overloading) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Composite Video Encoder 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo NTSC Equation (170M) 11/14/2018 Naohiro Shichijo, The University of Tokyo

NTSC composite signal spectrum 11/14/2018 Naohiro Shichijo, The University of Tokyo

Compositing waveforms 11/14/2018 Naohiro Shichijo, The University of Tokyo

Horizontal Blanking Interval 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Field Structure Even field Complete Frame Odd field 11/14/2018 Naohiro Shichijo, The University of Tokyo

Horizontal Blank Structure 11/14/2018 Naohiro Shichijo, The University of Tokyo

Introduction to Digital Video

Naohiro Shichijo, The University of Tokyo Signal Sampling Continuous analog signal is sampled in equally T interval. Nyquist’s theorem: Sampling frequency should be larger than twice of sampled signal frequency Sampling frequency: 4fsc … 14.3MHz (NTSC), 17.7MHz (PAL) (sampling frequency) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Aliasing 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Sampled spectrum Ideal spectrum Spectrum of filtered signal 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Quantization 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Digital Video Signals Composite Digital Video Component Digital Video 11/14/2018 Naohiro Shichijo, The University of Tokyo

Composite Digital Signal 4fsc NTSC, PAL composite digital Sampled at 4 times of sub carrier frequency ( fsc ) 8bits / 10bits per sample Used in D-2, D-3 etc. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo I-Q representation 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Sampling timings 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Quantization 10bits/sample 8 bits/sample (two LSBs are set zero, shifted twice) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Quantization in 244M 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Raster structures (NTSC) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Component Digital Signal CCIR601 standard. (ITU-R. 601) Several standard sampling rates: 4:2:2 sampling Luma at 13.5MHz, Chroma at 6.75MHz (2x3.375MHz) 4:1:1 sampling Luma at 13.5MHz (4x3.375MHz), Chroma at 3.375MHz 4:2:0 sampling Luma at 13.5MHz, Chroma at 6.75MHz (interleaved) 4:4:4 sampling Luma and chroma are sampled at 13.5MHz 11/14/2018 Naohiro Shichijo, The University of Tokyo

Sampling structure (4:2:2) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Horizontal Sync Relationship 11/14/2018 Naohiro Shichijo, The University of Tokyo

Multiplexing luma and chroma Total sample number is 1716 words (NTSC), 1728 words (PAL) per line. In digital active lines, 720 samples for Y, 360 samples for Cb and Cr  1440 samples in total. Digital blanking is 276 words (NTSC), 288 words (PAL) Samples are transferred in {Cb, Y, Cr} sequence 11/14/2018 Naohiro Shichijo, The University of Tokyo

Timing reference signals In component digital, analog sync signals are not sampled. 1440~1443 words: EAV (End of active video) 1712~1715 words: SAV (Start of active video) SAV and EAV is represented as “3FF 000 000 XYZ” XYZ: {(MSB)0,H,V,F,P0,P1,P2,P3,0,0(LSB)}2 F: field ID, V: vertical blanking information, H: horizontal blanking information, P0~P3: error correction bits. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Multiplex Structure 11/14/2018 Naohiro Shichijo, The University of Tokyo ITU-R BT.656-4

Naohiro Shichijo, The University of Tokyo Ancillary data spaces 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Ancillary data Data between EAV and SAV, and VLI are defined as “ancillary data.” Those are: Horizontal ancillary area: HANC Vertical ancillary area: VANC They start with sequence of “000 3FF 3FF.” HANC: 140,700 words/frame … 42.2Mb/s VANC: 56,160 words/frame … 13.5Mb/s In total … 55.7Mb/s (in NTSC) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Serial digital transmission Uses one coax cable. Multiplexes multi-channel audio and Video. Audio channels are inserted in ancillary areas. AES/EBU 2ch-digital audio … 3.072Mb/s In 4fsc NTSC format, 4 channels (or two 2-channels) can be multiplexed. SMPTE 259M-1997 Level A -- 143 Mb/s, NTSC Level B -- 177 Mb/s, PAL Level C -- 270 Mb/s, 525/625 component Level D -- 360 Mb/s, 525/625 component 11/14/2018 Naohiro Shichijo, The University of Tokyo

Serial Digital Interface (SDI) Channel coding NRZI (to reduce DC component compared to NRZ) Uses scrambler to further reduce DC component. G1(X) = X9 + X4 + 1. G2(X) = X + 1 Synchronizing TRS-ID (Timing reference signal and line number identification) TRS: 4 words (3FF 000 000 000) are inserted at 790~793th words in NTSC ID: 1word inserted at 794th word. Composed with line number and field number 11/14/2018 Naohiro Shichijo, The University of Tokyo

Serial Data Transport Interface (SDTI) Compatible with SDI Header in HANC User Data in Digital Active Video region (SAV~EAV) Can contain both fixed and variable length data packets in User Data region Usually, data packets will carry compressed video data E.g. MPEG2 (in D-10), DV (in D-7) etc... 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo SDTI Payload 11/14/2018 Naohiro Shichijo, The University of Tokyo

HDTV

Naohiro Shichijo, The University of Tokyo HDTV formats SMPTE 274M SMPTE 293M 11/14/2018 Naohiro Shichijo, The University of Tokyo

Basic difference with SDTV Primary Colors Luma and Chroma equations 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Raster Structure Progressive Systems  straightforward Interlace Systems  rather compilicated that is: Since active line is 1125, one frame is decomposed to 563 and 562 line fields. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Raster Structure of 1125P/I Systems 11/14/2018 Naohiro Shichijo, The University of Tokyo

HDTV Vertical Blanking Structure 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo HD-SDI and HD-SDTI HD-SDI: Almost same as SD versions. Same channel encoding Bit rate is 1.486Gb/sec Uses coax cable or optical fiber HD-SDTI: For compressed formats 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo HD-SDI C/Y payloads Each C/Y multiplexed line is treated as two lines. Original multiplexed line contains two headers. C then Y Double rate operation: doubles line numbers for progressive mode. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Interleaved Data Stream in HD-SDI 11/14/2018 Naohiro Shichijo, The University of Tokyo

SDTI Header and User Data Formats 11/14/2018 Naohiro Shichijo, The University of Tokyo

Digital Video Recording

Digital video tape recorders (DVTRs) SMPTE standardizes TV studio level DVTRs. D-1, D-2, D-3, D-5, HD-D5, D-6, D-7, D-9, D-10, D-11, D-12 Digital Betacam, Betacam SX are not SMPTE standard, but commonly used. 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Uncompressed DVTRs D-1, D-5: component digital D-2, D-3: composite digital D-6: component digital HDTV 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Compressed DVTRs D-7 (DVCPRO): 25Mb/s, 50Mb/s In 25Mb/s, 4:1:1, In 50Mb/s, 4:2:2 Compressed 1/3.3~1/5 by SMPTE314M (DV) D-9 (Digital S) 50Mb/s, 4:2:2 sampled, 1/3.3 compressed by DCT Betacam SX 4:2:2 sampled, 1/10 compressed by MPEG-2 (MP@ML) 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Compressed HDTV DVTRs HD-D5 Compressed 1/4.3 by DCT HDCAM 11/14/2018 Naohiro Shichijo, The University of Tokyo

Naohiro Shichijo, The University of Tokyo Related information SMPTE (professional video standards) http://www.smpte.org/ ITU (e.g. CCIR-601) http://www.itu.int/ IEC (mainly for consumer standards) http://www.iec.ch/ ATSC (ATV in US) http://www.atsc.org/ DVB (digital broadcast in Europe) http://www.dvb.org/ 11/14/2018 Naohiro Shichijo, The University of Tokyo