92 4 Digital Video Signal According to ITU-BT.R.601 (CCIR 601)
The luminance bandwidth is then limited to 5.75 MHz using a low-pass filter. The two color difference signals are limited to 2.75 MHz, i.e. the color resolution is clearly reduced compared with the brightness resolu- tion. This principle is familiar from children’s books where the impression of sharpness is simply conveyed by printed black lines. In analog televi- sion (NTSC, PAL, SECAM), too, the color resolution is reduced to about 1.3 MHz. The low-pass-filtered Y, CB and CR signals are then sampled and digitized by means of analog/digital converters. The A/D converter in the luminance branch operates at a sampling frequency of 13.5 MHz and the two CB and CR color difference signals are sampled at 6.75 MHz each.
Fig. 4.2. Sampling of the components in accordance with ITU-BT.R601
This meets the requirements of the sampling theorem: There are no more signal components above half the sampling frequency. The three A/D converters can all have a resolution of 8 or 10 bits. With a resolution of 10 bits, this will result in a gross data rate of 270 Mbit/s which is suitable for distribution in the studio but much too high for TV transmission via exist- ing channels (terrestrial, satellite or cable). The samples of all three A/D converters are multiplexed in the following order: CB Y CR Y CB Y ... In this digital video signal (Fig. 4.1.), the luminance value thus alternates with a CB value or a CR value and there are twice as many Y values as there are CB or CR values. This is called a 4:2:2 resolution, compared with the resolution immediately after the matrixing, which was the same for all components, namely 4:4:4.
SAV SAV
EAV EAV
Cb CrY Y Cb Y . . . .
Active video Blanking
Y
CB
CR
SAV = Start of active video EAV = End of active video 13.5 MHz luminance sampling frequency
6.75 MHz chrominance sampling frequency
Fig. 4.3. SAV and EAV code words in the ITU-BT.R601 signal
This digital signal can be present in parallel form at a 25 pin sub-D con- nector or serially at a 75 Ohm BNC socket. The serial interface is called SDI which stands for serial digital interface and has become the most widely used interface because a conventional 75-Ohm BNC cable can be used.
Within the data stream, the start and the end of the active video signal is marked by special code words called SAV (start of active video) and EAV (end of active video), naturally enough (Fig. 4.2.). Between EAV and SAV, there is the horizontal blanking interval which does not contain any information related to the video signal, i.e. the digital signal does not con- tain the sync pulse. In the horizontal blanking interval, supplementary in- formation can be transmitted such as, e.g. audio signals (embedded audio).
The SAV and EAV code words (Fig. 4.3.) consist of four 8 or 10 bit code words each. SAV and EAV begins with one code word in which all bits are set to one, followed by two words in which all bits are set to zero.
The fourth code word contains information about the respective field or the vertical blanking interval, respectively. This fourth code word is used for detecting the start of a frame, field and active picture area in the vertical direction. The most significant bit of the fourth code word is always 1. The next bit (bit 8 in a 10 bit transmission or bit 6 in an 8 bit transmission) flags the field; if this bit is set to zero, it is a line of the first field and if it is set to one, it is a line of the second field. The next bit (bit 7 in a 10 bit
TRS = Timing Reference Sequence 4 code words (SAV or EAV
= Start of active video or end of active video)
11111111(11) = 255/1023 00000000(00) = 0/0 1 F V H P3 P2 P1 P0 00
00000000(00) = 0/0
1 2 3 4
. . . . .
. . . . .
94 4 Digital Video Signal According to ITU-BT.R.601 (CCIR 601)
transmission or bit 5 in an 8 bit transmission) flags the active video area in the vertical direction. If this bit is set to zero, then this is the visible active video area and if not, it is the vertical blanking interval. Bit 6 (10 bit) or bit 4 (8 bit) provides information about whether the present code word is an SAV or an EAV. It is SAV if this bit is set to zero and EAV if it is not.
Bits 5...2 (10 bit) or 3...0 (8 bit) are used for error protection of the SAV and EAV code words. Code word 4 of the timing reference sequence (TRS) contains the following information:
F = Field (0 = 1st field, 1 = 2nd field)
V = Vertical blanking (1 = vertical blanking interval active)
H = SAV/EAV identification (0 = SAV, 1 = EAV)
P0, P1, P2, P3 = Protection bits (Hamming code)
Neither the luminance signal (Y) nor the color difference signals (CB, CR) use the full dynamic range available for them. There is a prohibited range which is reserved as headroom, on the one hand, and, on the other hand, allows SAV and EAV to be easily identified. A Y signal ranges be- tween 16 and 64 decimal (8 bits) or 240 and 960 decimal (10 bits).
Fig. 4.4. Level diagram
The dynamic range of CB and CR is 16 to 240 decimal (8 bits) or 64 to 960 decimal (10 bits). The area outside this range is used as headroom and for sync identification purposes.
0 mV 700 mV
0 mV
-350 mV 350 mV 255/1023
0 16/64 235/940
0 255/1023
128/512
16/64 240/960
Y C
B/C
RThis video signal conforming to ITU-BT.R601, which is normally available as an SDI (Serial Digital Interface) signal, forms the input signal to an MPEG SDTV encoder.
Physically the SDI signal is scrambled and NRZI encoded (NRZI = Non-Return-to-Zero code Inverted). The SDI spectrum is a sin(x)/x func- tion with its first zero at 270 MHz.
Bibliography: [ITU601], [MÄUSL4], [GRUNWALD]