Digital television
The transmission of television signals involves an analogue carrier wave- form that is modulated by the video (and audio) information. While the carrier is analogue, the video information may be analogue (analogue television) or digital known as digital television (DTV). In analogue television, the totality of the composite video, blanking and sync is transmitted in its original analogue format. In DTV, the video and audio information are first converted into a digital format composed of a series of zeros and ones (bits). The series of bits is then used to modulate an analogue carrier before broadcasting via an aerial. At the receiving end, the digitised video and audio information is converted back to their original analogue formats for viewing and listening by the user.
Broadcasting a high volume of information requires a very wide band- width which for analogue television is between 5 and 6 MHz. For digital video broadcasting (DVB), a bandwidth of 10 or more times wider is nec- essary. For this reason, data compression techniques are used to reduce the bandwidth to manageable proportions. In fact, data compression is so effective that more than one programme is made to fit within the band- width allocated for a single analogue channel. This is just one advantage of DTV broadcasting. Here are some more:
● Very good picture quality
● Increased number of programmes mentioned above
● Lower transmission power – reduces adjacent channel interference
● Lower signal-to-noise ratio
● No ghosting
Principles of digital video broadcasting
Broadcasting of DTV signals involves three steps as illustrated in Figure 3.1:
● Digitisation
● Compression
● Channel encoding
Digitisation is the process of converting the analogue video and audio sig- nals into a series of bits using an analogue to digital converter (ADC). To reduce the bandwidth requirements, data compression is used for both the video and audio information. This is carried out by the video and audio MPEG encoder which produces a series of video and audio packets known as the packetised elementary stream (PES). These are further broken up
into smaller 188-byte packets. Packets belonging to a number of different programmes are fed into a multiplexer to produce what is known as a transport stream (TS). Following the addition of error correction data by the FEC (forwards error correction) processor, the TS is used to modulate a carrier ready for broadcasting. Satellite DVB makes use of frequencies between 10,700 and 12,750 MHz using quadrature frequency shift keying (QFSK) while terrestrial DVB is restricted to the UHF band of frequecies currently available for analogue TV broadcasting with a channel band- width of 8 MHz in the UK and 6 MHz in the USA. A DVB multiplex occu- pying one analogue channel spectrum can accommodate from 3 to 10 different TV ‘channels’ or programmes.
Digitising the TV picture
Digitising a TV picture means sampling the contents of a picture frame by frame and, scan-line by scan-line (Figure 3.2). In order to maintain the quality of the picture, there must be at least as many samples per line as there are pixels, with each sample representing one pixel. For DTV, the pic- ture frame is a matrix of pixels: horizontal and vertical. The total number of pixels is the product of horizontal pixels (pixels/line) X vertical pixels (number of lines). The number of pixels will depend on the format used, e.g. standard definition television (SDTV; PAL or NTSC) or high definition television (HDTV). Let’s start with SDTV.
As we know from Chapter 2, PAL standard television uses 625 lines of which 576 are ‘active’ in that they may be used to carry video information and NTSC uses 525 lines system with 480 active lines. As for the number of pixels per line, SDTV specifies 720 pixels per line for both systems, giv- ing a total number of pixels per picture of
Each scan line will therefore be represented by 720 samples, and each sample will represent one pixel. Sample 1 represents pixel 1, sample 2 represents pixel 2, etc. The process is repeated for the second line, and so on until the end of the frame and then repeated all over again for the next frame. To ensure that the samples are taken at exactly the same point of the frame, the sampling frequency must be locked to the line frequency 15.625 kHz for PAL and 15.734 kHz for NTSC. For this reason, the sampling rate must be wholly divisible by either line frequency.