Digital Television Mpeg1 Mpeg2 And Principles Of The Dvb System 2nd Edition Herv Benot

Digital Television Mpeg1 Mpeg2 And Principles Of
The Dvb System 2nd Edition Herv Benot download
https://ebookbell.com/product/digital-television-mpeg1-mpeg2-and-
principles-of-the-dvb-system-2nd-edition-herv-benot-2330036
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Digital Television In Europe 1st Edition Wendy Van Den Broeck Jo
Pierson
https://ebookbell.com/product/digital-television-in-europe-1st-
edition-wendy-van-den-broeck-jo-pierson-51376524
Digital Television Systems Fundamentals Broadcast Systems Stefan Mozar
https://ebookbell.com/product/digital-television-systems-fundamentals-
broadcast-systems-stefan-mozar-54865254
Digital Television A Practical Guide For Engineers Dipling Fh Walter
Fischer Auth
https://ebookbell.com/product/digital-television-a-practical-guide-
for-engineers-dipling-fh-walter-fischer-auth-4189978
Digital Television Strategies Business Challenges And Opportunities
Alan Griffiths
https://ebookbell.com/product/digital-television-strategies-business-
challenges-and-opportunities-alan-griffiths-922616

Digital Television Technology And Standards John F Arnold Michael R
Frater
https://ebookbell.com/product/digital-television-technology-and-
standards-john-f-arnold-michael-r-frater-1076166
Digital Television Systems 1st Edition Marcelo S Alencar
https://ebookbell.com/product/digital-television-systems-1st-edition-
marcelo-s-alencar-1185754
Digital Television Production A Handbook Jeremy Orlebar
https://ebookbell.com/product/digital-television-production-a-
handbook-jeremy-orlebar-1219842
Cinematic Digital Television Chris Comerford
https://ebookbell.com/product/cinematic-digital-television-chris-
comerford-46521756
Understanding Digital Television An Introduction To Dvb Systems With
Satellite Cable Broadband And Terrestrial Tv Distribution 1st Edition
Larsingemar Lundstrom
https://ebookbell.com/product/understanding-digital-television-an-
introduction-to-dvb-systems-with-satellite-cable-broadband-and-
terrestrial-tv-distribution-1st-edition-larsingemar-lundstrom-2007278

This Page Intentionally Left Blank

Digital Television
MPEG-1, MPEG-2
and principles of the
DVB system
Second edition
Herve
Â Benoit

Focal Press
An imprint of Elsevier Science
Linacre House, Jordan Hill, Oxford OX2 8DP
225 Wildwood Avenue, Woburn MA 01801-2041
This book is adapted and translated by the author from Benoit, H.,
1996: La TeÂleÂvision NumeÂrique: MPEG-1, MPEG-2, et les principles
du systeÁme europeÂen DVB. Paris: Dunod
First published in Great Britain 1997 by Arnold
Second edition 2002
Copyright # 1997, 2002, Herve
Â Benoit. All rights reserved
The right of Herve
Â Benoit to be identified as the author of this work
has been asserted in accordance with the Copyright, Designs and
Patents Act 1988
No part of this publication may be reproduced in any material form (including
photocopying or storing in any medium by electronic means and whether
or not transiently or incidentally to some other use of this publication) without
the written permission of the copyright holder except in accordance with the
provisions of the Copyright, Designs and Patents Act 1988 or under the terms of
a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road,
London, England W1T 4LP. Applications for the copyright holder's written
permission to reproduce any part of this publication should be addressed
to the publisher
British Library Cataloguing in Publication Data
Benoit, Herve
Â
Digital television: MPEG-1, MPEG-2 and principles of the
DVB system. ± 2nd ed.
1. Digital television
I. Title
621.3 0
88
Library of Congress Cataloguing in Publication Data
A catalogue record for this book is available from the Library of Congress
ISBN 0 240 51695 8
For information on all Focal Press publications visit our website at:
www.focalpress.com
Typeset by Integra Software Services Pvt. Ltd, Pondicherry, India
Printed and bound in Great Britain

Contents
Preface ix
Acknowledgements xi
Introduction 1
1. Colour television: a review of current standards 3
1.1 Monochrome TV basics 3
1.2 Black and white compatible colour systems 6
2. Digitization of video signals 15
2.1 Why digitize video signals? 15
2.2 Digitization formats 15
2.3 Transport problems 21
3. Source coding: compression of video and audio signals 25
3.1 Some general data compression principles 25
3.2 Compression applied to images: the discrete
cosine transform (DCT) 28
3.3 Compression of fixed pictures 31
3.4 Compression of moving pictures (MPEG) 34
3.5 Compression of audio signals 48
4. Source multiplexing 58
4.1 Organization of the MPEG-1 multiplex: system
layer 58
4.2 Organization of the MPEG-2 multiplex:
program and transport streams 62
5. Scrambling and conditional access 74
5.1 Principles of the scrambling system in the
DVB standard 75
5.2 Conditional access mechanisms 77
5.3 Main conditional access systems 80

6. Channel coding (forward error correction) 81
6.1 Energy dispersal (randomizing) 82
6.2 Reed±Solomon coding (outer coding) 83
6.3 Forney convolutional interleaving (temporal
spreading of errors) 84
6.4 Convolutional coding (inner coding) 86
7. Modulation by digital signals 89
7.1 General discussion on the modulation of a
carrier by digital signals 90
7.2 Quadrature modulations 92
7.3 Modulation characteristics for cable and satellite
digital TV broadcasting (DVB-C and DVB-S) 95
7.4 OFDM modulation for terrestrial digital TV (DVB-T) 99
7.5 Summary of DVB transmission characteristics
(cable, satellite, terrestrial) 108
8. Reception of digital TV signals 110
8.1 Global view of the transmission/reception process 110
8.2 Composition of the integrated receiver decoder (IRD) 112
9. Middleware and interoperability aspects 124
9.1 Main proprietary middlewares used in Europe 127
9.2 The open European `middlewares' 131
10. Evolution: state of the art and perspectives 136
10.1 Digital terrestrial television 136
10.2 Evolution of the set-top box 138
10.3 New architectures 141
Appendix A: error detection and correction in digital
transmissions 146
A1.1 An error detecting code: the parity bit 146
A1.2 Block error correction codes 147
A1.3 Convolutional coding 151
Appendix B: spectral efficiency of cable and satellite
transmissions with DVB parameters 153
Appendix C: reception of the digital TV channels
of ASTRA and EUTELSAT 156
Appendix D: the main other digital TV systems 159
D1.1 The DSS system (satellite, USA) 159
D2.1 The ATSC system (terrestrial, USA) 160
D3.1 The ISDB-T system (terrestrial, Japan) 162
vi Contents

Appendix E: the IEEE1394 high speed
serial AV interconnection bus 165
Appendix F: the DiSEqC bus for antenna
system control 168
F1.1 The DiSEqC levels 168
F2.1 DiSEqC basic principles 169
F3.1 Different fields of the DiSEqC message 171
Appendix G: the common interface (DVB-CI) 175
Appendix H: example chipset for DVB decoders 178
Glossary of abbreviations, words and expressions 179
Abbreviations 179
Words and expressions 186
Bibliography 192
Books 192
Official documents (in English) 192
Some useful Internet addresses 193
Index 195
Contents vii

Preface
This book does not aim to make the reader an expert in digital
television (which the author himself is not). Rather, its purpose is
to describe and explain, as simply and as completely as possible, the
various aspects of the very complex problems that had to be solved
in order to define reliable standards for broadcasting digital pictures
to the consumer, and the solutions chosen for the European DVB
system (Digital Video Broadcasting) based on the international
MPEG-2 compression standard.
Thebookisintendedforreaderswithabackgroundinelectronicsand
some knowledge of conventional analogue television (a reminder of the
basic principles of existing television standards is presented for those
who require it) and for those with a basic digital background. The main
goal is to enable readers to understand the principles of this new
technology, to have a relatively global perspective on it, and, if they
wish, to investigate further any particular aspect by reading more
specialized and more detailed books. At the end, there is a short
bibliography and a glossary of abbreviations and expressions, which
will help readers to access some of these references.
For ease of understanding, after a general presentation of the
problem, the order in which the main aspects of digital television
broadcast standards are described follows the logical progression of
the signal processing steps on the transmitter side ± from raw
digitization used in TV studios to source coding (MPEG-2
compression and multiplexing), and on to channel coding (from
forward error correction to RF modulation). JPEG and MPEG-1
`predecessor' standards of MPEG-2 are also described, as MPEG-2
uses the same basic principles.
The book ends with a functional description of a digital IRD
(integrated receiver decoder), or set-top box, which the concepts
discussed in preceding chapters will help to demystify, and with a
discussion of future prospects.

This new edition includes an important update of the parts
concerning digital terrestrial television and interactive television
by cable, which were still at the end of the standardization process
at the time of the previous edition, as well as the new functionalities
of the post-2000 receivers and their evolution. A new chapter has
been added to cover the aspects of software interoperability, which
is a prerequisite to the generalization of digital television, especially
terrestrial, in view of analogue switch-off around the year 2010.
H. Benoit
x Preface

Acknowledgements
I would like to thank all those who lent me their support in the
realization of this book, especially:
. Philips Semiconductors labs for their training and many of the
figures illustrating this book;
. The DVB Project Office, the EBU Technical Publication Service
and the ETSI Infocentre for permission to reproduce the figures
of which they are the source.

Introduction
At the end of the 1980s, the possibility of broadcasting fully digital
pictures to the consumer was still seen as a faraway prospect, and
one that was definitely not technically or economically realistic
before the turn of the century. The main reason for this was the
very high bit-rate required for the transmission of digitized 525- or
625-line live video pictures (from 108 to 270 Mb/s without
compression). Another reason was that, at that time, it seemed
more urgent and important ± at least in the eyes of some politicians
and technocrats ± to improve the quality of the TV picture, and
huge amounts of money were invested by the three main world
players (first Japan, then Europe, and finally the USA) in order to
develop Improved Definition TeleVision (IDTV) and High Defin-
ition TeleVision systems (HDTV), with vertical resolutions from
750 lines for IDTV to 1125 or 1250 lines for HDTV.
Simply digitized, HDTV pictures would have required bit-rates
that were four times higher than `conventional' pictures, of the
order of up to one gigabit per second! This is why most of the
HDTV proposals (MUSE in Japan, HD-MAC in Europe, and
the first American HD proposals) were at that time defined as
analogue systems with a digital assistance which can be seen
as a prelude to fully digital compression.
However, by the beginning of the 1990s, the situation had
completely changed. Very quick development of efficient compres-
sion algorithms, resulting in, among other things, the JPEG
standard for fixed images and later the MPEG standard for moving
pictures, showed the possibility to reduce drastically the amount of
data required for the transmission of digital pictures (bit-rates from
1.5 to 30 Mb/s depending on the resolution chosen and the picture
content).

At the same time, continuous progress in IC technology allowed
the realization, at an affordable price, of the complex chips and
associated memory required for decompression of the digital
pictures. In addition, it appeared that the price of a HDTV receiver
would not quickly reach a level affordable by most consumers, not
so much due to the electronics cost, but mainly because of the very
high cost of the display, regardless of the technology used (big 16/9
tube, LCD projector or any other known technology). Further-
more, most consumers seemed more interested in the content and
the number of programmes offered than in an improvement in the
picture quality, and economic crises in most countries resulted in a
demand for `brown goods' oriented more towards the cheaper end
of the market.
Mainly on the initiative of the US industry, which could take
advantage of its traditional predominance in digital data proces-
sing to regain influence in the electronic consumer goods market,
studies have been reoriented towards the definition of systems
allowing diffusion of digital pictures with equivalent or slightly
better quality than current analogue standards, but with many
other features made possible by complete digitization of the signal.
The first digital TV broadcasting for the consumer started in mid-
1994 with the `DirecTV' project, and its success was immediate,
resulting in more than one million subscribers after one year.
However, the Europeans had not gone to sleep ± they decided at
the end of 1991 to stop working on analogue HDTV (HD-MAC),
and created the European Launching Group (ELG) in order to
define and standardize a digital TV broadcasting system. This gave
birth in 1993 to the DVB project (Digital Video Broadcasting),
based on the `main profile at main level' (MP@ML) of the inter-
national MPEG-2 compression standard.
MPEG-2 is downwards compatible with MPEG-1 and has pro-
visions for a compatible evolution towards HDTV by using higher
levels and profiles. This resulted in the standardization of three
variants for the various transmission media ± satellite (DVB-S),
cable (DVB-C) and terrestrial (DVB-T) ± which occurred between
1994 and 1996.
2 Introduction

1 Colour television: a
review of current
standards
Let us begin with a bit of history . . .
1.1 Monochrome TV basics
It should be borne in mind that all current TV standards in use
today are derived from the `black and white' TV standards started
in the 1940s and 1950s, which have defined their framework.
The first attempts at electromechanical television began at the
end of the 1920s, using the Nipkow disk for analysis and repro-
duction of the scene to be televised, with a definition of 30 lines and
12.5 images per second. This low definition resulted in a video
bandwidth of less than 10 kHz, allowing these pictures to be broad-
cast on an ordinary AM/MW or LW transmitter. The resolution
soon improved to 60, 90 and 120 lines and then stabilized for a
while on 180 lines (Germany, France) or 240 lines (England, USA)
around 1935. Scanning was `progressive', which meant that all lines
of the pictures were scanned sequentially in one frame, as depicted
in Fig. 1.1 (numbered here for a 625-line system).
These definitions, used for the first `regular' broadcasts, were the
practical limit for the Nipkow disk used for picture analysis;
the cathode ray tube (CRT) started to be used for display at the
receiver side. In order to avoid disturbances due to electromagnetic
radiation from transformers or a ripple in the power supply, the
picture rate (or frame rate) was derived from the mains frequency.
This resulted in refresh rates of 25 pictures/s in Europe and

30 pictures/s in the USA. The bandwidth required was of the order
of 1 MHz, which implied the use of VHF frequencies (in the order
of 40±50 MHz) for transmission. However, the spatial resolution of
these first TV pictures was still insufficient, and they were affected
by a very annoying flicker due to the fact that their refresh rate was
too low.
During the years just preceding World War II, image analysis
had become fully electronic with the invention of the iconoscope,
and definitions in use attained 405 lines (England) to 441 lines
(USA, Germany) or 455 lines (France), thanks to the use of inter-
laced scanning. This ingenious method, invented in 1927, consisted
of scanning a first field made of the odd lines of the frame and then
a second field made of the even lines (see Fig. 1.2), allowing the
picture refresh rate for a given vertical resolution to be doubled (50
or 60 Hz instead of 25 or 30 Hz) without increasing the bandwidth
required for broadcasting.
The need to maintain a link between picture rate and mains
frequency, however, inevitably led to different standards on both
sides of the Atlantic, even when the number of lines was identical
(as in the case of the 441-line US and German systems). Never-
theless, these systems shared the following common features:
. a unique composite picture signal combining video, blanking
and synchronization information (abbreviated to VBS, also
described as video baseband signal; see Fig. 1.3);
One frame of 625 lines
(575 visible)
Frame retrace
(50 lines)
1
2
3
4
5
6
570
571
572
573
574
575
Fig. 1.1 Schematic representation of progressive scanning
4 Colour television: a review of current standards

. an interlaced scanning (order 2), recognized as the best
compromise between flicker and the required bandwidth.
Soon afterwards, due to the increase in the size of the picture
tube, and taking into account the eye's resolution in normal view-
ing conditions, the spatial resolution of these systems still appeared
insufficient, and most experts proposed a vertical definition of
between 500 and 700 lines. The following characteristics were
finally chosen in 1941 for the US monochrome system, which
later became NTSC when it was upgraded to colour in 1952:
Two fields of 312.5 lines each
(2x287.5 visible)
First field retrace
(25 lines)
336
337
338
339
620
621
622
623
Second field retrace
(25 lines)
23
24
25
27
308
309
310
26
Fig. 1.2 Schematic representation of interlaced scanning (625 lines)
White
Black
level
Horizontal
synchronization
Horizontal
suppression
Synchronization
level
Visible part
Total line duration
Fig. 1.3 View of a line of a composite monochrome video signal
Monochrome TV basics 5

. 525 lines, interlaced scanning (two fields of 262.5 lines);
. field frequency, 60 Hz (changed to 59.94 Hz upon the
introduction of colour; see Note 1.1);
. line frequency, 15 750 Hz (60 262:5); later changed to 15 734
Hz with colour (59:94 262:5);
. video bandwidth, 4.2 MHz; negative video modulation;
. FM sound with carrier 4.5 MHz above the picture carrier.
After World War II, from 1949 onwards, most European
countries (except France and Great Britain) adopted the German
`GERBER' standard, also known as CCIR. It can be seen as an
adaptation of the US system to a 50 Hz field frequency, keeping a
line frequency as near as possible to 15 750 Hz; this allowed some
advantage to be taken of the American experience with receiver
technology. This choice implied an increased number of lines (ap-
proximately in the ratio 60/50) and, consequently, a wider band-
width in order to obtain well balanced horizontal and vertical
resolutions. The following characteristics were defined:
. 625 lines, interlaced scanning (two fields of 312.5 lines);
. field frequency, 50 Hz;
. line frequency, 15 625 Hz (50 312:5);
. video bandwidth, 5.0 MHz; negative video modulation;
. FM sound carrier 5.5 MHz above the picture carrier.
This has formed the basis of all the European colour standards
defined later (PAL, SECAM, D2-MAC, PAL ‡).
Until the beginning of the 1980s, different systems have been in
use in the UK (405 lines, launched in 1937 and restarted after a
long interruption during the war) and in France (819 lines,
launched in 1949 by Henri de France, who also invented the
SECAM system in 1957). These systems were not adapted to
colour TV for consumer broadcasting due to the near impossibility
of colour standard conversion with the technical means available at
that time, and were finally abandoned after a period of simulcast
with the new colour standard.
1.2 Black and white compatible colour systems
As early as the late 1940s, US TV set manufacturers and broad-
casting companies competed in order to define the specifications of

a colour TV system. The proposal officially approved in 1952 by
the FCC (Federal Communications Commission), known as NTSC
(National Television Standard Committee), was the RCA pro-
posal. It was the only one built on the basis of bi-directional compat-
ibility with the existing monochrome standard. A monochrome
receiver was able to display the new colour broadcasts in black
and white, and a colour receiver could, in the same way, display the
existing black and white broadcasts, which comprised the vast
majority of transmissions until the mid-1960s.
In Europe, official colour broadcasts started more than 10 years
later, in 1967, with SECAM (se
Â quentiel couleur a
Á me
Â moire) and
PAL (phase alternating line) systems.
Extensive preliminary studies on colour perception and a great
deal of ingenuity were required to define these standards which,
despite their imperfections, still satisfy most of the end users more
than 40 years after the first of them, NTSC, came into being. The
triple red/green/blue (RGB) signals delivered by the TV camera
had to be transformed into a signal which, on the one hand, could
be displayable without major artefacts on current black and white
receivers, and on the other hand could be transmitted in the band-
width of an existing TV channel ± definitely not a simple task.
The basic idea was to transform, by a linear combination, the
three (R, G, B) signals into three other equivalent components,
Y, Cb, Cr (or Y, U, V):
Y ˆ 0:587G ‡ 0:299R ‡ 0:114B is called the luminance signal
Cb ˆ 0:564 B Y† or U ˆ 0:493 (B Y) is called the blue
chrominance or colour difference
Cr ˆ 0:713 (R Y) or V ˆ 0:877 (R Y) is called the red
chrominance or colour difference
The combination used for the luminance (or `luma') signal has been
chosen to be as similar as possible to the output signal of a mono-
chrome camera, which allows the black and white receiver to treat
it as a normal monochrome signal. The two chrominance
(or `chroma') signals represent the `coloration' of the monochrome
picture carried by the Y signal, and allow, by linear recombination
with Y, the retrieval of the original RGB signals in the colour
receiver.
Studies on visual perception have shown that the human eye's
resolution is less acute for colour than for luminance transients.
This means, for natural pictures at least, that chrominance signals
can tolerate a strongly reduced bandwidth (one-half to one-quarter
of the luminance bandwidth), which will prove very useful for
Black and white compatible colour systems 7

putting the chrominance signals within the existing video spectrum.
The Y, Cb, Cr combination is the common point to all colour TV
systems, including the newest digital standards, which seems to
prove that the choices of the colour TV pioneers were not so bad!
In order to be able to transport these three signals in an existing
TV channel (6 MHz in the USA, 7 or 8 MHz in Europe), a
subcarrier was added within the video spectrum, modulated by
the reduced bandwidth chrominance signals, thus giving a new
composite signal called the CVBS (Colour Video Baseband Signal;
see Fig. 1.4). In order not to disturb the luminance and the black
and white receivers, this carrier had to be placed in the highest part
of the video spectrum and had to stay within the limits of the
existing video bandwidth (4.2 MHz in the USA, 5±6 MHz in
Europe; see Fig. 1.5).
Up to this point, no major differences between the three world
standards (NTSC, PAL, SECAM) have been highlighted. The
differences that do exist mainly concern the way of modulating
this subcarrier and its frequency.
1.2.1 NTSC
This system uses a line-locked subcarrier at 3.579 545 MHz
(ˆ 455 Fh=2), amplitude modulated with a suppressed carrier
following two orthogonal axes (quadrature amplitude modulation,
or QAM), by two signals, I (in phase) and Q (quadrature), carrying
the chrominance information. These signals are two linear combina-
tions of (R Y) and (B Y), corresponding to a 33
rotation of
the vectors relative to the (B Y) axis. This process results in a
White
Black
level
Synchro 4.7 s
μ
Suppression
Synchronization
level
52 s
μ
64 s
μ
12 s
μ
1.0 V
0.3 V
0 V
Burst
Fig. 1.4 View of a line of composite colour video signal (PAL or NTSC)

vector (Fig. 1.6), the phase of which represents the tint, and the
amplitude of which represents colour intensity (saturation).
A reference burst at 3.579 545 MHz with a 180
phase relative to
the B Y axis superimposed on the back porch allows the
receiver to rebuild the subcarrier required to demodulate I and Q
signals. The choice for the subcarrier of an odd multiple of half the
Amplitude
Subcarrier
chrominance
Sound
carrier
Chrominance
0 1 2 3 4 5
4.43 5.5
f (Mhz)
Fig. 1.5 Frequency spectrum of the PAL signal
I ( =123°)
ϕ
+ ( – )
R Y
Red Magenta
IM
QM
Q ( =33°)
ϕ
Burst ( =180°)
ϕ
Yellow
Blue
+ ( – )
B Y
Green Cyan
S
a
t
u
r
a
t
i
o
n
α=Tint
Fig. 1.6 Colour plan of the NTSC system

line frequency is such that the luminance spectrum (made up of
discrete stripes centred on multiples of the line frequency) and the
chrominance spectrum (discrete stripes centred on odd multiples
of half the line frequency) are interlaced, making an almost perfect
separation theoretically possible by the use of comb filters in the
receiver.
Practice, however, soon showed that NTSC was very sensitive
to phase rotations introduced by the transmission channel, which
resulted in very important tint errors, especially in the region
of flesh tones (thus leading to the necessity of a tint correction
button accessible to the user on the receivers and to the famous
`never twice the same colour' expression). This led Europeans to
look for solutions to this problem, which resulted in the SECAM
and PAL systems.
1.2.2 SECAM
This standard eliminates the main drawback of the NTSC
system by using frequency modulation for the subcarrier, which is
insensitive to phase rotations; however, FM does not allow
simultaneous modulation of the subcarrier by two signals, as does
QAM.
The clever means of circumventing this problem consisted of
considering that the colour information of two consecutive lines
was sufficiently similar to be considered as identical. This
reduces chroma resolution by a factor of 2 in the vertical direc-
tion, making it more consistent with the horizontal resolution
resulting from bandwidth reduction of the chroma signals.
Therefore, it is possible to transmit alternately one chrominance
component, D0
b ˆ 1:5(B Y), on one line and the other,
D0
r ˆ 1:9(R Y), on the next line. It is then up to the receiver
to recover the two D0
b and D0
r signals simultaneously, which can
be done by means of a 64 ms delay line (one line duration) and
a permutator circuit. Subcarrier frequencies chosen are
4.250 MHz (ˆ 272 Fh) for the line carrying D0
b and 4.406250
MHz (ˆ 282 Fh) for D0
r.
This system is very robust, and gives a very accurate tint
reproduction, but it has some drawbacks due to the frequency
modulation ± the subcarrier is always present, even in non-
coloured parts of the pictures, making it more visible than in
NTSC or PAL on black and white, and the continuous nature
of the FM spectrum does not allow an efficient comb filtering;
rendition of sharp transients between highly saturated colours is
not optimum due to the necessary truncation of maximum FM

deviation. In addition, direct mixing of two or more SECAM
signals is not possible.
1.2.3 PAL
This is a close relative of the NTSC system, whose main drawback
it corrects. It uses a line-locked subcarrier at 4.433 619 MHz
(ˆ 1135=4 ‡ 1=625 Fh), which is QAM modulated by the two
colour difference signals U ˆ 0:493 (B Y) and V ˆ 0:877
(R Y).
In order to avoid drawbacks due to phase rotations, the phase of
the V carrier is inverted every second line, which allows cancella-
tion of phase rotations in the receiver by adding the V signal from
two consecutive lines by means of a 64 ms delay line (using the same
assumption as in SECAM, that two consecutive lines can be
considered as identical). In order to synchronize the V demodu-
lator, the phase of the reference burst is alternated from line to line
between ‡135
and 135
compared to the U vector (0
).
Other features of PAL are very similar to NTSC. In addition to
the main PAL standard (sometimes called PAL B/G), there are two
other less well known variants used in South America in order to
accommodate the 6 MHz channels taken from NTSC:
. PAL M used in Brazil (525 lines/59.94 Hz, subcarrier at 3.575
611 MHz);
. PAL N used in Argentina (625 lines/50 Hz, subcarrier at 3.582
056 MHz).
1.2.4 MAC (multiplexed analogue components)
During the 1980s, Europeans attempted to define a common
standard for satellite broadcasts, with the goal of improving picture
and sound quality by eliminating drawbacks of composite systems
(cross-colour, cross-luminance, reduced bandwidth) and by using
digital sound. This resulted in the MAC systems, with a compatible
extension towards HDTV (called HD-MAC).
D2-MAC is the most well known of these hybrid systems, even if
it did not achieve its expected success, due to its late introduction
and an earlier development of digital TV than anticipated. It
replaces frequency division multiplexing of luminance, chromin-
ance and sound (bandwidth sharing) of composite standards by a
time division multiplexing (time sharing). It is designed to be

compatible with normal (4/3) and wide screen (16/9) formats and
can be considered in some aspects as an intermediate step on the
route to all-digital TV signal transmission.
On the transmitter side, after sampling (Note 1.2) and analogue-
to-digital conversion, Y, Cb and Cr signals are time-compressed by
a factor of 2/3 for Y and 1/3 for Cb and Cr, scrambled if required,
and then reconverted into analogue form in order to be transmitted
sequentially over one line duration (see Fig. 1.7 illustrating one line
of a D2-MAC signal). The part of the line usually occupied by
synchronization and blanking is replaced by a burst of so-called
duobinary data (hence the `D2' in D2-MAC). These data carry the
digital sound, synchronization and other information such as tele-
text, captioning and picture format (4/3 or 16/9), and in addition,
for pay TV programmes, they carry the access control messages of
the Eurocrypt system used with D2-MAC.
As in SECAM, Cb and Cr chroma components are transmitted
alternately from line to line in order to reduce the necessary band-
width and obtain equivalent resolutions along the two axes of the
picture for the chrominance. This resolution corresponds to the
so-called 4:2:0 format (see Section 2.2.2, p. 19); it is almost equiva-
lent to the professional 4:2:2 format used in TV studios. Time
division multiplexing results in the total elimination of cross-colour
and cross-luminance effects, and in a luminance bandwidth of
5 MHz, a substantial improvement compared with PAL or SECAM.
1.2.5 PAL ‡
This is a recent development, the primary objective of which was to
allow terrestrial transmission of improved definition 16/9 pictures
0.5 V
0.4 V
0.4 V
0.5 V
0.5 V
0.5 V
0.5 V
Reference
level
Clamp
period
Sound and
data
Chrominance
( or )
U V
Luminance
Y
64 s
μ
Fig. 1.7 Composition of a line of a D2-MAC signal

(on appropriate receivers) in a compatible way with existing 4/3
PAL receivers (Note 1.3). To do this, the PAL‡ encoder trans-
forms the 576 useful lines of a 16/9 picture into a 4/3 picture in
letterbox format (a format often used for the transmission of films
on TV, with two horizontal black stripes above and below the
picture). The visible part occupies only 432 lines (576 3=4) on
a 4/3 receiver, and additional information for the PAL‡ receiver
is encoded in the remaining 144 lines.
The 432-line letterbox picture is obtained by vertical low-pass
filtering of the original 576 lines, and the complementary high-pass
filtering is transmitted on the 4.43 MHz subcarrier during the
144 black lines, which permits the PAL‡ receiver to reconstruct
a full-screen 16/9 high resolution picture.
In order to obtain the maximum bandwidth for luminance
(5 MHz) and to reduce cross-colour and cross-luminance, the
phase of the subcarrier of the two interlaced lines of consecutive
fields is reversed. This process, known as `colorplus', allows
(by means of a frame memory in the receiver) cancellation of
cross-luminance by adding the high part of the spectrum of two
consecutive frames, and reduction of cross-colour by subtracting
them.
A movement compensation is required to avoid artefacts intro-
duced by the colorplus process on fast moving objects, which,
added to the need for a frame memory, contributes to the relatively
high cost of current PAL‡ receivers. The PAL‡ system results in
a subjective quality equivalent to D2-MAC on a 16/9 receiver
in good reception conditions (high signal/noise ratio).
PAL
burst
0.5 V
Video
Clock reference
+14 information bits
27.4 s
μ
44.5 s
μ 19.5 s
μ
11 s
μ
Fig. 1.8 View of line 23 of a PAL‡ signal (WSS bits)

In order to inform the receiver of the format of the programme
being broadcast (4/3 or 16/9), signalling bits (WSS: wide screen
signalling) and additional information (sound mode, etc.) are
added to the first half of line 23 (Fig. 1.8), which permits the
receiver to adapt its display format. The WSS signal can also be
used by ordinary PAL 16/9 receivers simply to modify the vertical
amplitude according to the format, which is sometimes referred to
as the `poor man's PAL‡ '.
After this introduction (hopefully not too lengthy), we will
now attempt to describe as simply as possible the principles
which have allowed the establishment of new all-digital television
standards and services, the impact of which is still difficult to
comprehend.
Note 1.1
This slight change in line and field frequencies was introduced in
order to minimize the visual effect of beat frequency between
sound (4.50 MHz) and colour (3.58 MHz) subcarriers in the
receiver. This change was done by using the sound intercarrier
as a reference for the line frequency
15 734 ˆ 4 500 000=286†
Note 1.2
D2-MAC is based on the 4:2:0 digital format (720 points/line for
Y and 360 for Cb and Cr), but for practical reasons, these
numbers had to be slightly reduced to 700 and 350, respectively.
This is due to the fact that the duration of 720 samples at 13.5
MHz (53.33 ms) is more than the useful part of the analogue
video line (52 ms), which could disturb clamping circuits in the
receiver.
Note 1.3
PAL‡ development took place between 1990 and 1992; after a
period of experimental transmissions, official broadcasts started
in Germany and other countries after the international 1995
Berlin radio/TV exhibition (IFA). PAL‡ is officially adopted
by most countries currently using the PAL system. The WSS
format signalling information will also be used independently of
PAL‡ for conventional PAL or SECAM transmissions.

2 Digitization of video
signals
2.1 Why digitize video signals?
For a number of years, video professionals at television studios
have been using various digital formats, such as D1 (components)
and D2 (composite), for recording and editing video signals. In
order to ease the interoperability of equipment and international
programme exchange, the former CCIR (Comite
Â Consultatif Inter-
national des Radiocommuncations; Note 2.1) has standardized
conditions of digitization (recommendation CCIR-601) and inter-
facing (recommendation CCIR-656) of digital video signals in
component form (Y, Cr, Cb in 4:2:2 format).
The main advantages of these digital formats are that they allow
multiple copies to be made without any degradation in quality, and
the creation of special effects not otherwise possible in analogue
format, and they simplify editing of all kinds, as well as permitting
international exchange independent of the broadcast standard to
be used for diffusion (NTSC, PAL, SECAM, D2-MAC, MPEG).
However, the drawback is the very important bit-rate, which makes
these formats unsuitable for transmission to the end user without
prior signal compression.
2.2 Digitization formats
If one wants to digitize an analogue signal of bandwidth Fmax, it is
necessary to sample its value with a sampling frequency Fs of at

least twice the maximum frequency of this signal to keep its integ-
rity (Shannon sampling theorem). This is to avoid the negative
aliasing effects of spectrum fall-back: in effect, sampling a signal
creates two parasitic sidebands above and below the sampling
frequency, which range from Fs Fmax to Fs ‡ Fmax, as well as
around harmomics of the sampling frequency (Fig. 2.1).
In order to avoid mixing of the input signal spectrum and the
lower part of the first parasitic sideband, the necessary and suffi-
cient condition is that Fs Fmax Fmax, which is realized if
Fs 2Fmax. This means that the signal to be digitized needs to be
efficiently filtered in order to ensure that its bandwidth does not
exceed Fmax ˆ Fs=2.
For component video signals from a studio source, which can
have a bandwidth of up to 6 MHz, the CCIR prescribes a sampling
frequency of Fs ˆ 13:5 MHz locked on the line frequency (Note
2.2). This frequency is independent of the scanning standard, and
represents 864 Fh for 625-line systems and 858 Fh for 525-line
systems. The number of active samples per line is 720 in both cases.
In such a line-locked sampling system, samples are at the same
fixed place on all lines in a frame, and also from frame to frame,
and so are situated on a rectangular grid. For this reason, this
sampling method is called orthogonal sampling (Fig. 2.2), as op-
posed to other sampling schemes used for composite video sam-
pling (4 Fsc subcarrier locked sampling for instance).
The most economic method in terms of bit-rate for video signal
digitization seems, a priori, to be to use the composite signal as a
source; however, the quality will be limited by its composite nature.
Taking into account the fact that 8 bits (corresponding to 256
quantization steps) is the minimum required for a good signal to
quantization noise ratio (Sv=Nq 59 dB; Note 2.3), the bit-rate
Amplitude
4.5 MHz
F = 10 MHz
s
4.5 MHz 4.5 MHz
0 1 2 3 4 5 10 15
f (MHz)
Fig. 2.1 Spectrum of a sampled signal (when Fs 2 Fmax)
16 Digitization of video signals

required by this composite digitization is 13:5 8 ˆ 108 Mb/s,
which is already a lot!
However, digitization of a composite signal has little advantage
over its analogue form for production purposes (practically the
only one is the possibility of multiple copies without degradation).
This is why this is not the preferred method for source signal
digitization in broadcast applications, as the composite signal is
not very suitable for most signal manipulations (editing, compres-
sion) or international exchanges.
2.2.1 The 4:2:2 format
Recommendation CCIR-601, established in 1982, defines digitiza-
tion parameters for video signals in component form based on a
Y, Cb, Cr signal in 4:2:2 format (four Y samples for two Cb samples
and two Cr samples) with 8 bits per sample (with a provision for
extension to 10 bits per sample). The sampling frequency is 13.5
MHz for luminance and 6.75 MHz for chrominance, regardless of
the standard of the input signal. This results in 720 active video
samples per line for luminance, and 360 active samples per line for
1
2
3
Pictures
4
Pixels
Lines
Fig. 2.2 Orthogonal sampling structure of a picture
Digitization formats 17

each chrominance. The position of the chrominance samples cor-
responds to the odd samples of the luminance (see Fig. 2.3).
Chrominance signals Cr and Cb being simultaneously available
at every line, vertical resolution for chrominance is the same as for
luminance (480 lines for 525-line systems, 576 lines for 625-line
systems). The total bit-rate resulting from this process is
13:5 8 ‡ 2 6:75 8 ˆ 216 Mb/s. With a quantization on 10
bits, the bit-rate becomes 270 Mb/s! However, if one takes into
account the redundancy involved in digitizing the inactive part of
the video signal (horizontal and vertical blanking periods), the
useful bit-rate goes down to 166 Mb/s with 8 bits per sample.
These horizontal and vertical blanking periods can be filled with
other useful data, such as digital sound, sync and other informa-
tion.
Recommendation CCIR-656 defines standardized electrical in-
terfacing conditions for 4:2:2 signals digitized according to recom-
mendation CCIR-601. This is the format used for interfacing D1
digital video recorders, and is therefore sometimes referred to as
the D1 format.
The parallel version of this recommendation provides the signal
in a multiplexed form (Cr1, Y1, Cb1, Y2, Cr3, Y3, Cb3 . . . ) on an 8-bit
parallel interface, together with a 27 MHz clock (one clock period
per sample). Synchronization and other data are included in the
data flow. The normalized connector is a DB25 plug.
There is also a serial form of the CCIR-656 interface for
transmission on a 75 coaxial cable with BNC connectors,
Luminance
Chrominance
4:2:2
Fig. 2.3 Position of samples in the 4:2:2 format

requiring a slightly higher bit-rate (243 Mb/s) due to the use of 9
bits per sample in this mode.
2.2.2 4:2:0, SIF, CIF and QCIF formats
For applications that are less demanding in terms of resolution,
and in view of the bit-rate reduction, a certain number of by-
products of the 4:2:2 format have been defined, as follows.
The 4:2:0 format
This format is obtained from the 4:2:2 format by using the same
chroma samples for two successive lines, in order to reduce the
amount of memory required in processing circuitry while at
the same time giving a vertical resolution of the same order as
the horizontal resolution. Luminance and horizontal chrominance
resolutions are the same as for the 4:2:2 format, and thus
. luminance resolution: 720 576 (625 lines) or 720 480 (525
lines);
. chrominance resolution: 360 288 (625 lines) or 360 240 (525
lines).
Figure 2.4 shows the position of chroma samples in the 4:2:0
format.
In order to avoid the chrominance line flickering observed in
SECAM at sharp horizontal transients (due to the fact that one
Luminance
Chrominance
4:2:0
Fig. 2.4 Position of samples in the 4:2:0 format
Digitization formats 19

chrominance comes from the current line and the second comes
from the preceding one), Cb and Cr samples are obtained by inter-
polating 4:2:2 samples of the two successive lines they will `colour-
ize' at display time.
This 4:2:0 format is of special importance as it is the input format
used for D2-MAC and MPEG-2 (MP@ML) coding.
The SIF format (source intermediate format)
This format is obtained by halving the spatial resolution in both
directions as well as the temporal resolution, which becomes 25 Hz
for 625-line systems and 29.97 Hz for 525-line systems. Depending
on the originating standard, the spatial resolutions are then:
. luminance resolution: 360 288 (625 lines) or 360 240 (525
lines);
. chrominance resolution: 180 144 (625 lines) or 180 120 (525
lines).
Figure 2.5 illustrates the position of the samples in the SIF format.
Horizontal resolution is obtained by filtering and subsampling the
input signal. The reduction in temporal and vertical resolution is
normally obtained by interpolating samples of the odd and even
fields, but is sometimes achieved by simply dropping every second
field of the interlaced input format. The resolution obtained is the
Luminance
Chrominance
SIF
Discarded samples
Fig. 2.5 Position of samples in the SIF format

base for MPEG-1 encoding, and is giving a so-called `VHS-like'
quality in terms of resolution.
The CIF format (common intermediate format)
This is a compromise between European and American SIF for-
mats: spatial resolution is taken from the 625-line SIF (360 288)
and temporal resolution from the 525-line SIF (29.97 Hz). It is the
basis used for video conferencing.
The QCIF format (quarter CIF)
Once again, this reduces the spatial resolution by 4 (2 in each
direction) and the temporal resolution by 2 or 4 (15 or 7.5 Hz). It
is the input format used for ISDN videotelephony using the H261
compression algorithm.
2.3 Transport problems
It is clear that a bit-rate of the order of 200 Mb/s, as required by the
4:2:2 format, cannot be used for direct broadcast to the end user, as
it would occupy a bandwidth of the order of 40 MHz with a
64-QAM modulation (6 bits/symbol) used for cable, or 135 MHz
with a QPSK modulation (2 bits/symbol) used for satellite. This
would represent 5±6 times the bandwidth required for transmission
of an analogue PAL or SECAM signal, and does not even take into
account any error correction algorithm (these concepts will be
explained later in Chapters 6 and 7 on channel coding and modu-
lation).
Compression algorithms, however, have been in use for some
years for contribution links in the field of professional video, which
reduce this bit-rate to 34 Mb/s, but this is still too high for con-
sumer applications, as it does not give any advantage in terms of
capacity over existing analogue transmissions. It was the belief that
this problem would not be solved economically in the foreseeable
future (in large part due to the cost of the memory size required)
that gave birth in the 1980s to hybrid standards such as D2-MAC
(analogue video, digital sound) and delayed the introduction of
100% digital video. However, the very rapid progress made in
compression techniques and IC technology in the second half
of the 1980s made these systems obsolete soon after their
introduction.
Transport problems 21

The essential conditions required to start digital television
broadcast services were the development of technically and eco-
nomically viable solutions to problems which can be classified into
two main categories:
. Source coding ± this is the technical term for compression. It
encompasses all video and audio compression techniques used to
reduce as much as possible the bit-rate (in terms of Mb/s required
to transmit moving pictures of a given resolution and the associated
sound) with the lowest perceptible degradation in quality.
. Channel coding ± this consists of developing powerful error
correction algorithms associated with the most spectrally
efficient modulation techniques (in terms of Mb/s per MHz),
taking into account the available bandwidth and the foreseeable
imperfections of the transmission channel.
Taking into account the fact that many programmes can be
transmitted on one RF channel, the sequence of operations to be
performed on the transmitter side is roughly as illustrated in Fig.
2.6. We will follow the order of the functional boxes in this figure
when we discuss them in the following chapters.
MPEG
coding
MPEG
coding
Multiplexing
(+ scrambling)
Forward error
correction (FEC)
Filtering and
modulation
Up-conversion
and amplification
Programme 1
Programme n
Source coding + multiplexing Channel coding + modulation
Broadcasting
network
Fig. 2.6 Sequence of main operations on the broadcasting side
Note 2.1
The CCIR was the radiocommunication branch of the former
CCITT (Comite
Â Consultatif International du Te
Â le
Â graphe et du
Te
Â le
Â phone), recently renamed the ITU (International Telecom-
munications Union). The CCIR is now called the ITU-R.
Note 2.2
For some multimedia applications (video in PCs mainly), other
sampling frequencies are often used in order to obtain so-called

square pixels to ease mixing of live video and computer images
without aspect ratio distortion. In fact, the aspect ratio of a
standard TV picture and of a computer display is 4/3 (ratio of
the width to the height of the picture); in order to have identical
resolutions in the horizontal and vertical directions, which is the
case for today's computer display formats, it is necessary that
the ratio of the number of pixels per line to the number of useful
lines is 4/3. This is not the case for CCIR-601 derived formats,
neither in 625-line standards (720=576 4=3) nor in 525-line
standards (720=480 4=3).
For 525-line standards (480 useful lines), square pixels imply
480 4=3 ˆ 640 pixels/line, which is obtained with a sampling
frequency of 12.2727 MHz. This is not an accident since this
resolution of 640 480 corresponds to the basic VGA graphics
mode ± this mode is, in fact, an uninterlaced or progressive
variant of the NTSC scanning standard (line frequency ˆ
31 468 Hz, frame frequency ˆ 59.94 Hz). For 625-line standards
(576 useful lines), square pixels imply 576 4=3 ˆ 768 pixels/
line, which requires a sampling frequency of 14.75 MHz.
Note 2.3
Dynamic
The dynamic D of a signal with a maximum peak-to-peak ampli-
tude VPP, quantized with m steps (with m ˆ 2b
, where b is the
number of bits of the quantization), is the ratio between VPP and
the maximum peak value of the quantization error, which is
equal to the quantization step Q.
By definition, Q is equal to the maximum peak-to-peak
amplitude VPP divided by the number of quantization steps m,
i.e. Q ˆ VPP/m ˆ VPP/2b
. Hence, the dynamic D (in dB) is equal
to
D dB† ˆ 20 log VPP=Q† ˆ 20 log VPP 2b
=VPP†
ˆ 20 log 2b
ˆ b 20 log 2
Hence
D b 6dB
Example 1 (video). A quantization with 8 bits (b ˆ 8) results in
D 48 dB.
Transport problems 23

Example 2 (audio). A quantization with 16 bits (b ˆ 16) results in
D 96 dB.
Signal to quantization noise ratio
If Q is the quantization step, the quantization noise voltage Nq is
equal to Q=

12
p
.
For a video signal, VPP is equal to the black-to-white ampli-
tude VBW, and so Q ˆ VBW=m ˆ VBW=2b
. The signal to quanti-
zation noise ratio, SV=Nq, is the ratio of the black-to-white
signal VBW to the quantization noise voltage Nq:
SV=Nq dB† ˆ 20 log VBW 2b

12
p
=VBW†
ˆ 20 log 2b

12
p
† b 6 ‡ 20 log

12
p
or
SV=Nq D ‡ 10:8dB
Therefore, in the case of example 1 above (video signal quan-
tized with 8 bits), D 48 dB and SV=Nq 58:8 dB.
For an audio signal, the signal to quantization noise ratio, SA/
Nq, is the ratio of the root mean square (RMS) signal VRMS to
the quantization noise voltage Nq. If we assume a sinusoidal
signal of maximum peak-to-peak amplitude VPP, the corres-
ponding maximum RMS voltage is VRMS ˆ VPP=2

2
p
. Thus
SA=Nq dB† ˆ 20 log VPP 2b

12
p
=VPP 2

2†
p
ˆ 20 log 2b

12
p
=2

2†
p
SA=Nq dB† b 6 ‡ 20 log

12
p
log 2

2†
p
D ‡ 20 0:09
or
SA=Nq D ‡ 1:8 dB
Thus, in the case of example 2 (audio signal quantized with 16
bits), D 96 dB and SA=Nq 97:8 dB.

3 Source coding:
compression of video
and audio signals
In the preceding chapter, we explained why compression was an
absolute must in order to be able to broadcast TV pictures in a
channel of acceptable width. A spectrum bandwidth comparable to
conventional analogue broadcasts (6±8 MHz for cable or terrestrial
broadcasts, 27±36 MHz for satellite) implies in practice
maximum bit-rates of the order of 30±40 Mb/s, with the necessary
error correction algorithms and modulation schemes, which are
explained in Chapters 6 and 7.
We will now examine the principles and various steps of video and
audio compression which allow these bit-rates (and in fact much less)
to be achieved, and which are currently being used in the various video/
audio compression standards. These compression methods use general
datacompressionalgorithmsapplicabletoanykindofdata,andexploit
the spatial redundancy (correlation of neighbouring points within an
image) and the specificities of visual perception (lack of sensitivity of
the eye to fine details) for fixed pictures (JPEG), and the very high
temporal redundancy between successive images in the case of moving
pictures (MPEG). In the same way, audio compression methods
exploit particularities of the human aural perception to reduce bit-
rates by eliminating inaudible information (psychoacoustic coding).
3.1 Some general data compression principles
3.1.1 Run length coding (RLC)
When an information source emits successive message elements
which can deliver relatively long series of identical elements

(which, as explained later in this chapter, is the case with the DCT
after thresholding and quantization), it is advantageous to transmit
the code of this element and the number of successive occurrences
rather than to repeat the code of the element; this gives a variable
compression factor (the longer the series, the bigger the compres-
sion factor). This type of coding which does not lose any informa-
tion is defined as reversible. This method is commonly employed
for file compression related to disk storage or transmission by
computers (zip etc.); it is also the method used in fax machines.
3.1.2 Variable length coding (VLC) or entropy coding
This bit-rate reduction method is based on the fact that the prob-
ability of occurrence of an element generated by a source and
coded on n bits is sometimes not the same (i.e. equiprobable) for
all elements among the 2n
different possibilities. This means that, in
order to reduce the bit-rate required to transmit the sequences
generated by the source, it is advantageous to encode the most
frequent elements with less than n bits and the less frequent ele-
ments with more bits, resulting in an average length that is less than
a fixed length of n bits.
However, if this is to be done in real time, it implies a previous
knowledge of the probability of occurrence of each possible ele-
ment generated by the source. We have this knowledge, for example,
in the case of the letters of the alphabet in a given language, and
this allows this method to be used for text compression. This
method is also valid for video images compressed by means of
DCT, where energy is concentrated on a relatively small number
of coefficients, as opposed to the temporal representation of the
video signal where all values are almost equiprobable.
One can demonstrate that the information quantity Q trans-
mitted by an element is equal to the logarithm (base 2) of the
inverse of its probability of appearance p:
Q ˆ log2 1=p† ˆ log2 p†
The sum of the information quantity of all elements generated by
a source multiplied by their probability of appearance is called the
entropy, H, of the source:
H ˆ
X
i
pi log2 1=pi†
26 Source coding: compression of video and audio signals

The goal of variable length coding (VLC), or entropy coding, is
to approach, as near as is possible, the entropic bit-rate (corres-
ponding to an averaged number of bits per element as near as
possible to the source's entropy). The most well-known method
for variable length coding is the Huffmann algorithm, which
assumes previous knowledge of the probability of each element.
It works in the following way (illustrated in Fig. 3.1):
. Each element is classified in order of decreasing probability,
forming an `occurrence table' (left part of Fig. 3.1).
. The two elements of lowest probability are then grouped into
one element, the probability of which is the sum of the two
probabilities. Bit `0' is attributed to the element of lowest
probability and `1' to the other element; this reduces by one the
number of elements to be classified.
. The new element is then grouped in the same way with the
element having the next highest probability. `0' and `1' are
attributed in the same way as above, and the process is
continued until all the elements have been coded (sum of the
probability of the last two elements ˆ 100%).
. In this way, the Huffmann coding tree is built (central part of Fig.
3.1): the code for each element is obtained by positioning
sequentially the bits encountered in moving along the Huffmann
tree from left to right.
Element p (%)
Huffmann coding tree
( ) = sum of the probability of the two elements in a column
xx Code
110100
A
C
G
E
H
B
F
D
40
20
14
10
06
05
03
02
0
111
101
100
1100
11011
110101
0
1
1
0
0
1
0
0
1
0
1
0
(100)
(60)
(24)
(36)
(16)
(10)
(5)
1
1
Fig. 3.1 Application of the Huffmann coding algorithm (grey zones indicate horizontal
links)
Some general data compression principles 27

To illustrate this method, we have assumed a source generating
eight elements with the following probabilities: p(A) ˆ 40%,
p(B) ˆ 50%, p(C) ˆ 20%, p(D) ˆ 2%, p(E) ˆ 10%, p(F) ˆ 3%,
p(G) ˆ 14%, p(H) ˆ 6%. In this example, the average word length
after coding (sum of the products of the number of bits of each
element and its probability) is 2.51 bits, while the entropy
H ˆ
P
i pi log2 1=pi† is equal to 2.44 bits; this is only 3% more
than the optimum, a very good efficiency for the Huffmann algo-
rithm. In this example with eight elements, a pure binary coding
would require 3 bits per elements, so the compression factor
achieved with the Huffmann coding is
ˆ 2:51=3:00 ˆ 83:7%
This type of coding is reversible (it does not lose information) and
can be applied to video signals as a complement to other methods
which generate elements of non-uniform probability (DCT fol-
lowed by quantization for instance). The overall gain can then be
much more important.
3.2 Compression applied to images: the discrete
cosine transform (DCT)
The discrete cosine transform is a particular case of the Fourier
transform applied to discrete (sampled) signals, which decomposes
a periodic signal into a series of sine and cosine harmonic func-
tions. The signal can then be represented by a series of coefficients
of each of these functions.
Without developing the mathematical details, we will simply
indicate that, under certain conditions, the DCT decomposes the
signal into only one series of harmonic cosine functions in phase
with the signal, which reduces by half the number of coefficients
necessary to describe the signal compared to a Fourier transform.
In the case of pictures, the original signal is a sampled bidimen-
sional signal, and so we will also have a bidimensional DCT
(horizontal and vertical directions), which will transform the lumi-
nance (or chrominance) discrete values of a block of N N pixels
into another block (or matrix) of N N coefficients representing
the amplitude of each of the cosine harmonic functions.
In the transformed block, coefficients on the horizontal axis
represent increasing horizontal frequencies from left to right, and
on the vertical axis they represent increasing vertical frequencies

from top to bottom. The first coefficient in the top left corner
(coordinates: 0, 0) represents null horizontal and vertical frequen-
cies, and is therefore called the DC coefficient, and the bottom
right coefficient represents the highest spatial frequency compon-
ent in the two directions.
In order to reduce the complexity of the circuitry and the
processing time required, the block size chosen is generally 8 8
pixels (Fig. 3.2), which the DCT transforms into a matrix of 8 8
coefficients (Fig. 3.3). A visual representation of the individual
contribution of each coefficient to the appearance of the original
block of 8 8 pixels can be seen in Fig. 3.4: the appearance of the
original picture block can be obtained by averaging each of the 64
squares in Fig. 3.4 by its coefficient and summing the results.
720 pixels (90 blocks)
576
pixels
(72
blocks)
Picture Block
8 pixels
143
131
147
151
118
121
127
129
150
134
144
143
124
93
142
117
136
126
134
132
112
116
119
103
114
143
136
139
134
136
147
125
120
134
113
115
126
124
162
122
112
123
111
116
129
137
149
109
129
124
140
134
132
138
123
128
134
133
129
131
125
131
132
132
8
pixels
Fig. 3.2 Cutting out blocks of 8 8 pixels (values represent the luminance of a pixel)
8x8 pixel block
143
131
147
151
118
121
127
129
150
134
144
143
124
93
142
117
136
126
134
132
112
116
119
103
114
143
136
139
134
136
147
125
120
134
113
115
126
124
162
122
112
123
111
116
129
137
149
109
129
124
140
134
132
138
123
128
134
133
129
131
125
131
132
132
Horizontal frequency
–11
–26
1033
–4
–8
0
–17
20
24
11
0
7
12
–4
–16
11
11
–9
11
1
20
10
–1
26
0
20
2
2
12
–2
14
–3
–2
8
8
3
37
–12
6
–1
–1
–7
–6
–3
4
31
8
–18
–3
4
–8
24
–10
5
–2
7
0
15
–3
3
6
–7
6
0
13
9
2
7
4
11
0
7
Matrix of 8 x 8 coefficients
0
1
2
3
4
5
6
7
Vertical
frequency
DCT
Fig. 3.3 Transformation of a block of 8 8 pixels into a matrix of 8 8 coefficients using
the DCT
Compression applied to images 29

Depending on the number of details contained in the original
block, the high frequency coefficients will be bigger or smaller, but
generally the amplitude decreases rather quickly with the fre-
quency, due to the smaller energy of high spatial frequencies in
most `natural' images. The DCT thus has the remarkable property
of concentrating the energy of the block on a relatively low number
of coefficients situated in the top left corner of the matrix. In
addition, these coefficients are decorrelated from each other.
These two properties will be used to advantage in the next steps
of the compression process.
Up to this point, there is no information loss: the DCT transform
process is reversible. However, due to the psycho-physiological
specificities of human vision (reduced sensitivity to high spatial
frequencies), it is possible, without perceptible degradation of
the picture quality, to eliminate the values below a certain
threshold function of the frequency. The eliminated values are
replaced by 0 (an operation known as thresholding); this part of the
Vertical
frequency
Fig. 3.4 Contribution of each of the DCT coefficients to the appearance of the 8 8 pixel
block

process is obviously not reversible, as some data are thrown
away. The remaining coefficients are then quantized with an accur-
acy decreasing with the increasing spatial frequencies, which
once again reduces the quantity of information required to encode
a block; here again the process is not reversible, but it has little
effect on the perceived picture quality. The thresholding/quantiza-
tion process is illustrated in Fig. 3.5.
The thresholding and quantization parameters can be used
dynamically to regulate the bit-rate required to transmit moving
pictures, as will be explained in Section 3.4.
A serial bitstream is obtained by `zig-zag' reading of the
coefficients, as shown in Fig. 3.6. This method is one of those
allowing a relatively long series of null coefficients to be obtained
as quickly as possible, in order to increase the efficiency of the
following steps ± run length coding followed by variable length
coding (see Section 3.1).
3.3 Compression of fixed pictures
The first applications aimed at reducing the amount of information
required for coding fixed pictures appeared in the 1980s, and they
had as their primary objective the significant reduction of the size
of graphics files and photographs in view of storing or transmitting
them. In 1990, the ISO (International Standards Organization)
created an international working group called JPEG (Joint Photo-
graphic Experts Group) which had the task of elaborating an
international compression standard for fixed pictures of various
resolutions in Y, Cr, Cb or RGB format. The resulting international
standard (widely known as JPEG) was published in 1993 under the
129
0
0
0
–1
1
–1
0
1
1
0
–1
0
0
0
0
Thresholding
+ quantization
1
1
0
0
1
0
1
0
1
–1
0
1
0
0
0
0
2
–1
0
0
0
0
0
0
1
0
–1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
–11
–26
1033
–4
–8
0
–17
20
24
11
7
12
–4
–16
11
11
–9
11
20
10
–1
26
0
20
2
12
–2
14
–3
–2
8
8
37
–12
6
–1
–1
–7
–6
–3
31
8
–18
–3
4
–8
24
–10
–2
7
0
15
–3
3
6
–7
0
13
9
2
7
4
11
0
Fig. 3.5 Result of thresholding and quantization
Compression of fixed pictures 31

reference ISO/IEC 10918, and it can be considered as a toolbox for
fixed picture compression. We will not describe it in detail, as it is
not the object of this book, but we will nevertheless go through its
main steps, as it has largely inspired the way in which MPEG
works.
It should be noted that JPEG compression can be either lossy or
lossless (reversible), depending on the application and the desired
compression factor. Most common applications use the lossy
method, which allows compression factors of more than 10 to be
achieved without noticeable picture quality degradation, depend-
ing on the picture content. We will only examine the case of lossy
JPEG compression, as the coding of I (intra) pictures of MPEG
uses the same process; lossless JPEG compression uses a different
predictive coding which is not based on DCT, so we will not discuss
it here.
Lossy JPEG compression can be described in six main steps:
1. Decomposition of the picture into blocks ± the picture, generally
in Y, Cb, Cr format, is divided into elementary blocks of 8 8
pixels (Fig. 3.2), which represents for a 4:2:2 CCIR-601 picture
a total number of 6480 luminance (Y) blocks and 3240 blocks
for each Cr and Cb component. Each block is made up of 64
numbers ranging from 0 to 255 (when digitized on 8 bits) for
luminance, and 128 to ‡127 for chrominance Cr and Cb.
0
1 2 3 4 5 6 7
1
2
3
4
5
6
7
0
Vertical
frequency
Last AC
coefficient
First AC
coefficient
DC coefficient
Fig. 3.6 `Zig-zag' reading of the coefficients of the matrix

2. Discrete cosine transform ± as explained previously, the DCT
applied to each Y, Cb, Cr block generates for each one a new
8 8 matrix made up of the coefficients of increasing spatial
frequency as one moves away from the origin (upper left corner)
which contains the DC component representing the average
luminance or chrominance of the block. The value of these
coefficients decreases quickly when going away from the origin
of the matrix, and the final values are generally a series of small
numbers or even zeros. So, if the block is of uniform luminance
or chrominance, only the DC coefficient will not be zero, and
only this coefficient will have to be transmitted.
3. Thresholding and quantization ± this step takes into account the
specificities of human vision, particularly the fact that the eye
does not distinguish fine details below a certain luminance level.
It consists of zeroing the coefficients below a predetermined
threshold, and quantizing the remaining ones with decreasing
accuracy as the frequency increases. Contrary to the 63 other
(AC) coefficients, the DC coefficient is DPCM coded (differ-
ential pulse code modulation) relative to the DC coefficient of
the previous block, which allows a more accurate coding with a
given number of bits. This allows the visibility of the blocks on
the reconstructed picture to be reduced, as the eye, although not
very sensitive to fine details, is nevertheless very sensitive to
small luminance differences on uniform zones.
4. Zig-zag scan ± except for the DC coefficient, which is treated
separately, the 63 AC coefficients are read using a zig-zag scan
(Fig. 3.6) in order to transform the matrix into a flow of data
best suited for the next coding steps (RLC/VLC).
5. Run length coding ± in order to make the best possible use of the
long series of zeros produced by the quantization and the zig-
zag scan, the number of occurrences of zero is coded, followed
by the next non-zero value, which reduces the amount of
information to transmit.
6. Variable length coding (Huffmann coding) ± this last step uses a
conversion table in order to encode the most frequently
occurring values with a short length, and the less frequent
values with a longer one. These last two steps (RLC and VLC)
alone ensure a compression factor of between 2 and 3.
When the compression/decompression time is not of prime
importance, which is often the case for fixed pictures, all the
above-described steps can be done entirely using software. There
Compression of fixed pictures 33

are, however, a number of specialized processors which can speed
up this process very much. The simplified principle of a JPEG
decoder can be seen in the block diagram in Fig. 3.7.
3.4 Compression of moving pictures (MPEG)
In 1990, the need to store and reproduce moving pictures and the
associated sound in digital format for multimedia applications on
various platforms led the ISO to form an expert group along the
same lines as JPEG, with members coming from the numerous
branches involved (computer industry, telecoms, consumer
electronics, semiconductors, broadcasters, universities, etc.); this
group was called MPEG (Motion Pictures Experts Group).
The first outcome of its work was the International Standard
ISO/IEC 11172, widely known as MPEG-1. The main goal was to
allow the storage on CD-ROM or CD-I (single speed at that time)
Compressed data stream
VLC coding table
Quantization table
VLC+RLC decoder
De-quantizer
Inverse
DCT
Decoded
picture
8x8 block
Fig. 3.7 Principle of JPEG decoding

of live video and stereo sound, which implied a maximum bit-rate of
1.5 Mb/s. In addition to the intrinsic spatial redundancy exploited by
JPEG for fixed pictures, coding of moving pictures allows exploit-
ation of the very important temporal redundancy between successive
pictures which make up a video sequence.
Given the very high compression rate objective for these
applications (more than 100 compared to the original 4:2:2/
CCIR 601 picture), the difficulty of the task is reduced from
the beginning of the process by sacrificing the resolution. The
format chosen for the pictures to be encoded is the SIF format
(described in Section 2.2.2 ± uninterlaced pictures, 360 288 @
25 Hz or 360 240 @ 30 Hz depending on the original video
standard), which corresponds roughly to the resolution of a
consumer video recorder.
The sound compression algorithm used for the accompanying
audio channels is known as MUSICAM, also used in the European
digital radio system (DAB, Digital Audio Broadcasting). We will
examine the principles of audio compression in Section 3.5.
The MPEG-1 standard consists of three distinct parts, published
in November 1992:
. MPEG-1 system (ISO/IEC 11172-1): defines the MPEG-1
multiplex structure (cf. Chapter 4);
. MPEG-1 video (ISO/IEC 13818-2): defines MPEG-2 video
coding;
. MPEG-1 audio (ISO/IEC 13818-3): defines MPEG-2 audio
coding.
However, the picture quality of MPEG-1 was not suitable for
broadcast applications, since, among other things, it did not take
into account the coding of interlaced pictures or evolution towards
HDTV. The MPEG group thus worked on the definition of a
flexible standard optimized for broadcasting. This international
standard is known as MPEG-2.
As its predecessor, MPEG-2 is specified in three distinct parts,
published in November 1994:
. MPEG-2 system (ISO/IEC 13818-1): defines the MPEG-2
streams (cf. Chapter 4);
. MPEG-2 video (ISO/IEC 13818-2): defines MPEG-2 video
coding;
. MPEG-2 audio (ISO/IEC 13818-3): defines MPEG-2 audio
coding.
Compression of moving pictures (MPEG) 35

MPEG-2 is, among other things, the source coding standard used
by the European DVB (Digital Video Broadcasting) TV broad-
casting system, which is the result of the work started in 1991 by the
ELG (European Launching Group), later to become the DVB
committee.
3.4.1 Principles behind the video coding of MPEG-1 (multimedia
applications)
As indicated previously, the main objective for MPEG-1 was to
reach a medium quality video with a constant total bit-rate of
1.5 Mb/s for storing video and audio on CD-ROM. The video
part uses 1.15 Mb/s, the remaining 350 kb/s being used by
audio and additional data required by the system and other
information. However, the MPEG-1 specification is very flex-
ible and allows different parameters to be chosen depending on
the compromise between encoder complexity, compression rate
and quality.
The video coding uses the same principles as lossy JPEG, to
which new techniques are added to form the MPEG-1 `toolbox';
these techniques exploit the strong correlation between successive
pictures in order to considerably reduce the amount of informa-
tion required to transmit or store them. These techniques, known
as `prediction with movement compensation', consist of deducing
most of the pictures of a sequence from preceding and even
subsequent pictures, with a minimum of additional information
representing the differences between pictures. This requires the
presence in the MPEG encoder of a movement estimator, which is
the most complex function and greatly determines the encoder's
performance; fortunately, this function is not required in the
decoder.
As we are talking about moving pictures, decoding has to be
accomplished in real time (this means an acceptable and constant
processing delay); this implies, for the time being at least, some
specialized hardware. The coding, which is much more complex,
can be done in more than one `pass' for applications where real
time is not required but where quality is of prime importance
(engraving of disks for instance); real time (which does not mean
null processing time) will, however, be required for many applica-
tions, such as `live' video transmissions.
The practical realization of the encoder is therefore a trade-off
between speed, compression rate, complexity and picture quality.
In addition, synchronization time and random access time to a
sequence have to be maintained within an acceptable limit (not

exceeding 0.5 s), which restricts the maximum number of pictures
that can be dependent on the first picture to between 10 and 12 for
a system operating at 25 pictures/s.
The different types of MPEG pictures
MPEG defines three types of pictures (Note 3.1) which are
arranged as shown in Fig. 3.8.
. I (intra) pictures are coded without reference to other pictures, in
a very similar manner to JPEG, which means that they contain
all the information necessary for their reconstruction by the
decoder; for this reason, they are the essential entry point for
access to a video sequence. The compression rate of I pictures is
relatively low, and is comparable to a JPEG coded picture of a
similar resolution.
. P (predicted) pictures are coded from the preceding I or P
picture, using the techniques of motion compensated prediction.
P pictures can be used as the basis for next predicted pictures,
but since motion compensation is not perfect, it is not possible to
extend very much the number of P pictures between two I
pictures. The compression rate of P pictures is significantly
higher than for I pictures.
. B (bidirectional or bidirectionally predicted) pictures are coded by
bidirectional interpolation between the I or P picture which
precedes and follows them. As they are not used for coding
B
I B P B P
Prediction Prediction
Interpolation Interpolation
Fig. 3.8 Concatenation of the three types of pictures in MPEG
Compression of moving pictures (MPEG) 37

Exploring the Variety of Random
Documents with Different Content

rocks; they would kill a stag in the adjoining forest; they would visit
the curé of Chevreuse, and astonish the good man,—a sport which
she by no means disliked: but while they were arranging all these
schemes on the open space before the chateau, a courier was seen
riding up from the gates, and when he came near he handed the
duchess two letters.
The blood left her cheek as she read, and, instantly drawing Edward
aside, she said, We must part at once. You go on as fast as possible
to Gray. Wait there two days, and, if you hear no more, ride forward
to Turin. As for myself, look here. And she put a paper into his
hand. It was a copy of the decree banishing her to Lorraine, there to
remain upon her own estates till the king's further pleasure.
Order your horses quickly, she said. Then come to my chamber
for the sum Montagu left for you. Glimpses of sunshine! glimpses of
sunshine in this April-day life! and then dark clouds and heavy
showers.
In an hour, Edward Langdale rode away from Dampierre. He was
grave and silent. What was in his heart who can tell? but he certainly
did not view the world more brightly, or feel more confidence in
human nature, than he had done before that short visit.

CHAPTER XXVI.
Edward Langdale rode on from place to place, sometimes quickly,
sometimes slowly, as the condition of the roads and the nature of
the country required; and, strangely enough for a journey in those
days, neither accident nor adventure befell him. One thing excited
his curiosity and suspicion, however. At Trapes, where he passed the
first night after leaving the house of Madame de Chevreuse, when
he had finished his supper and was just retiring to rest, he caught
for a moment, on the somewhat darksome stairs, one glance of a
face he thought he had seen before. He could not identify it, indeed,
for it was lost as soon as seen; but it instantly carried his mind back
to his adventure with the two Savoyards, and he felt almost sure
that face belonged to one of them. But neither of the two strangers
appeared the next morning; and Pierrot and Jacques both assured
him that their horses were not in the stable.
There are faces that haunt us both in night and daydreams; and
Edward was almost led to believe that one of these spectres of the
imagination had taken possession of him; for twice or three times
before he reached Gray that face again crossed him for a moment,
and always when no one else was present who could confirm or
remove his suspicions.
Those were not pleasant days to live in; and it is a very difficult
thing for any one born in and accustomed to the bad comfortable
modern days to realize those good old times. Espionage was then a
great science, an honorable profession, practised by great dignitaries
and men of high degree. Words brought men's heads to the block,
and thoughts often conducted to a prison. There was no need of
overt acts: intentions were quite sufficient; and friends and foes
were so continually changing places that no one could tell that the

thoughts uttered in the confidence of familiar intercourse would not
be brought forward a few days or weeks later to lead one to the
dungeon and the rack. Yet it is wonderful, unaccountable, how freely
and daringly men spoke their mind,—how the grave condemnation,
the witty lampoon, or the hideous libel, was disseminated without
ceremony. Men laughed and had their heads chopped off,—and
would have laughed still if they could have been fixed on again, I do
believe; for nothing seemed a warning or a restraint.
Edward, however, born in a country where neither the reign of the
Tudor nor of the Stuart had been able to crush out the spirit of
liberty, loved not to be watched; and there is always something
more alarming in the indefinite than the definite danger. He could
not divine what was the object of the two strangers, if, indeed, they
had any object, in thus persisting in following him. The cardinal had
lacked no opportunity of detaining him at Nantes, or of arresting him
on his journey, if he had thought fit; and yet he could not clear his
mind from suspicion till he reached Franche Comté and found
himself beyond the power of the French minister.
It may be necessary to remind the reader that Franche Comté was
not annexed to France till the year 1668; and at the time of which I
now write the important town of Gray was a fortified place,
consisting of the city on the high ground strongly walled, and a
suburb on the bank of the Saône, defended merely by a small
battery. For a long period of troublous times, so frequent had been
the visits of French exiles to Lorraine, Burgundy, and Franche Comté,
that safe-conducts or passports from one country to another were
very generally dispensed with in the country and in open towns; but
in fortresses some trouble was experienced; and it is probable that
the directions which the Duchesse de Chevreuse had given Edward
Langdale to stop in the faubourg were intended to guard against his
detention. The inn which she had named to him was good, however,
—perhaps better than that in the upper town; and the appointed
two days of Edward's stay passed dully but not unpleasantly. The
horses were refreshed and the two men none the worse for the

repose. For Edward himself, too, perhaps two days of thought were
beneficial. Every man, in the toil and tumult and hurry of the world,
requires some moment to pause and consider his position, to decide
upon his future course, to apply the lesson of past errors, to take
breath as it were amidst the bustle of existence. Edward was like a
stout swimmer who had been suddenly plunged into a torrent, and
was likely to be carried away by the flood which for the last three
months had been whirling confusedly round him; and those two
days at Gray were like a little island of dry ground where he could
rest and scan his way to the opposite bank, avoiding the rocks and
eddies which might impede or destroy him. It is a quaint old
proverb, but a true one, that a man who does not look clearly
before him will often have to look sadly behind him; and happy is
he who has both the will and the time to do so.
Those two days then with Edward passed in almost uninterrupted
thought; but at last the night of the second day came, and yet
neither message nor letter had arrived. Supper had been eaten, and
the horses had been ordered for daybreak on the following morning
to proceed to Turin, when, toward nine o'clock, the landlord brought
in a scrap of writing, asking Edward if that was intended for him. It
was addressed in English,—Master Edward Langdale,—and
underneath was written, Join me at Chambéry or Aix. I shall be
there from the twenty-ninth till the first.
No name was signed, but the writing was Lord Montagu's; and the
landlord, on being questioned, said the paper had been given to him
by a courier from Arnay le Duc going to Vesoul, who had gone on his
way as soon as he had left it.
Now, Edward's knowledge of geography was considerable, and, as
far as France and England were concerned, minute; but he had at
Gray got somewhat out of his latitude, and the landlord had to be
consulted as to the road to Aix and Chambéry. The good man was
learned upon the subject, however, knew every inch of the road, he
said, and could find his way in the dark. It was true, he added, that
it was rather a wild way, and carriages could hardly go one-half the

distance; but, as the gentleman had horses, it would be easily
managed. He must first go straight to Dole, then from Dole to Lons-
le-Saulnier, from Lons-le-Saulnier to Bourg or Nantua, and thence to
the Pont du Sault. After that, he said, came Bellay and Aix and
Chambéry; but there the traveller would have to ask every step of
his way. It was a five days' journey, he remarked, and, ride as hard
as you would, it would take four and a half.
Edward did ride hard, and the first part of the way was overcome in
a much shorter space of time than the good host had anticipated;
nor was it till the party had passed Bourg that any thing like
difficulties occurred. It is as pleasant a ride in fine weather as any
one can take, for the roads are now good and the scenery
exceedingly picturesque without being fatiguingly grand; but neither
Edward nor ourselves have any time to pause upon the beauties of
nature. The roads, however, were then in a very different condition
from that which they now display; and, indeed, the wonder-working
eighteenth and nineteenth centuries have done more for few
countries than for the districts lying between the Jura and the Rhone
and Saône.
On the twenty-seventh of July, Edward Langdale and his party were
within one short day's journey of Aix, and the early morning when
they set out was fresh and beautiful. The hot summer sun was
shaded by the rocks and forests, and the air was cooled by the
mountain-breeze. As he was earlier than the first of the days named
by Lord Montagu, the young traveller suffered his horses to proceed
leisurely. But in this he made a mistake. Man always wants more
money and time than he calculates upon; and nobody can tell what
the want of an hour or a guinea may bring about.
As every one knows, the country which Edward had now to traverse
is a land of rocks and mountains, of rivers and lakes. Not three miles
can be passed without encountering some stream or torrent hurrying
down to join the great Rhone; and at every mile, as the road then
went, was some steep ascent or descent, flanked with rugged cliffs,
sometimes covered with dark forests, sometimes naked and gray,

with immense masses of stone impending over the traveller's head
without the root of a single tree to bind them to the crag, while high
up in front the Mont du Chat was seen from time to time rearing its
rugged front and seeming to close the pass. About one o'clock, over
the edges of the hills some heavy clouds were seen rising, knotty
and dull, and of a deep lead-color, except where the sun tipped their
edges with an ochrey yellow. The wind was from the northeast, and
the clouds were coming from the south. But they did not heed the
breeze, which soon began to fail before them.
Let us ride faster, said Edward: the road is good here. And on he
went, keeping his eye on the heavy masses, but fearing no greater
inconvenience than a wetting. He had never travelled in Savoy
before. However, by quick trotting he saved himself and his followers
for about two hours; but by the end of that time the sun was hidden
and great drops began to fall. Then came the thunder echoing
through the hills, and then a complete deluge. Every thing turned
gray, and the old castles which strew that part of the country could
hardly be distinguished from the rocks on which they stood.
Two more hours were passed by the travellers under an overhanging
shelf of rock, which afforded some shelter, not only to themselves,
but also to their horses. But at the end of that time the rain had had
the effect of loosening some parts of the cliff, and several large
masses of stone began to fall, giving them warning to retreat as
soon as possible.
The thunder was now more distant and the flashes of lightning
farther apart; but the rain continued to fall, not so heavily, but in a
dull, incessant pour. There was nothing to be done but to ride on,
and, even then, but slow progress could be made; for the roads
were cut up in a terrible manner, the smaller streams were swollen
so as to be well nigh impassable, and here and there the way was
nearly blocked up by piles of rock and gravel. Night was rapidly
coming on; no human habitation was in sight except a scattered old
tower here and there, and that in ruins.

At length, just as the sun sank, a more formidable obstacle than
ever presented itself. Where the road took a rapid descent between
some high rocky ground on the right and the Rhone in flood upon
the left, just at the spot where one of the branches of the Guiers
joins the larger river, an immense mass of rock, undermined by the
torrent, had fallen across the mouth of the stream, which, thus
blocked up, had flooded the whole road. By the side of the water,
gazing disconsolately at the rushing and whirling current, was a
group of men, some four in number. It was too dark for Edward to
distinguish who they were at any distance, but when he came nearer
he perceived his two old friends the Savoyard blacksmiths, and two
laborers of the country, whom the fall of the rock and the
consequent inundation had, it seemed, cut off from their own
cottages on the other side.
Ah! bon jour, bon jour, seigneur! said one of the blacksmiths, who
had dismounted, and was holding his horse by the bridle: we came
all along the road with you, after all, but we kept out of your way for
fear of your pistols. Here is a pretty pass! We shall not get over to-
night, these men say.
Can we find no place of shelter this side? asked Edward, whose
suspicion of the two men had been greatly abated by finding they
had quietly pursued their way to Savoy. The blacksmith shook his
head.
I saw an old castle about half a mile back, said the young
Englishman: it was not far up the mountain.
All ruined! No roof, replied the other. Ask them yourself.
But Edward could not make either of the peasants comprehend a
word he said. We must do something, he remarked. It is growing
darker every moment, and it would give us some sort of covering,
were it but under an old arch. Hark! there are horses coming on the
other side. Those men will be into the torrent if they do not mind.

And, raising his voice, he shouted aloud to warn the horsemen, who
were dashing on at furious pace from the side of Aix.
The wind set the other way, and the roaring of the water was loud,
so that it is probable his shout was not heard, for the next moment
there was a plunge into the water and then a loud cry for help.
Edward sprang instantly from his horse and advanced to the very
verge of the stream.
For Heaven's sake, Master Ned, for Heaven's sake, do not try it!
cried Pierrot, catching his arm.
Here, take the horse, said Edward, sharply. Let go my arm.
A flash of lightning came at that moment, faint, indeed, but
sufficient to show him a horse carried away toward the Rhone, a
horseman who had pulled up just in time upon the other brink, and
a man struggling in the water and trying to hold by a smooth mass
of fallen rock, just in the middle of the torrent, about twelve yards
from him. He paused not to consider, but ran as far as he could up
the water, dashed in, and swam with all his strength toward the
drowning man, whom he could just distinguish. Borne down by the
current, he drifted right to the rock, calling aloud, in French, Do not
touch me, and I will save you!
Such warnings are usually vain. The man's first effort was to clutch
him; but Edward was prepared, and kept him off, catching him
tightly by the back of the neck. We have said that he was a good
and practised swimmer; but neither skill nor strength would probably
have carried him across that small space of twelve yards against that
powerful current. But Jacques Beaupré caught sight of him, and
exclaimed, Here, Pierrot, catch my hand. Let us all be drowned in
company. And, running in till the water reached his shoulders and
almost carried him off his feet, he contrived to grasp Edward's arm
and pull him on till he could touch ground.

The young lad was almost exhausted, for the man, of whom he had
never loosed his hold, had struggled to the last to grasp him, and
the few moments since he had left the rock had been all one
confused scene of strife amongst the dark and eddying waters.
Here; let me take him, sir, said Jacques: if ever a man's life was
nobly saved, it is his. And, throwing his brawny arms round the
stranger, who struggled still, he carried him on to the road.
Edward paused for a moment, as soon as he could resist the stream,
to draw breath, and then slowly joined the rest. They had laid the
stranger down on the bank, and for a moment or two he remained
quite still, though his panting breath showed that his life was in no
danger.
Here, moosoo, take some of this, said one of the blacksmiths,
pouring some spirit out of a bottle into the stranger's mouth: you
owe that young seigneur something; for if he had not been here you
would have been out of Savoy by this time.
I know it; I know it, said the rescued man, faintly. Where is he?
which is he?
Look! look! cried Pierrot: there is a light up there, in one, two,
three windows. That must be in the old chateau which these fellows
said was all in ruins. Let us go up. We shall none of us ever get dry
here, it is raining so hard.
Are you able, sir, to walk up to that castle? asked Edward,
speaking to the stranger, who had now raised himself upon his arm.
I fear your poor horse is lost beyond all hope.
Let the fiery brute go, said the other, petulantly: if he would have
obeyed the rein I should not have been in this plight. I will try to
accompany you in a moment. But what castle is that? It must be
Groslie, I think.
He did not speak very good French; but, calling to one of the
Savoyard peasants, he addressed him in his own language, of which

he seemed to have a perfect command.
The good man instantly began to speak fast and gesticulate
vehemently; and, translating as best he could the language of signs,
Edward concluded that the Savoyard was trying to dissuade the
gentleman from going to the old chateau he had seen.
What does he say? asked the young Englishman: he seems
unwilling we should go.
Oh, he talks nonsense, answered the stranger: he will have it that
the place is haunted, and says that no one is ever seen there by day,
but that those lights appear from time to time at night,—smugglers,
more likely, or coiners; but we are too many for them to do us any
harm. As he spoke he raised himself slowly upon his feet and said
to the friendly blacksmith, Give me some more of those strong
waters, my friend. I will pay you well for them.
The man readily supplied him, and he professed himself ready to
proceed; but the two peasants could not be induced by any means
to accompany the rest. One of the blacksmiths, however, produced a
lantern and candle from the packs which each carried behind his
saddle, and the party set out, not without fresh remonstrances from
the boors.
If they be devils, we do not fear them, replied the stranger, and
then added some directions which probably referred to the servant,
who had been able to stop his horse in time and remained on the
other side of the torrent.
The peasants seemed to treat the stranger with much respect; but
even when, by the aid of a flint and steel, the lantern was lighted, it
was impossible for Edward to discern more of the other's person
than sufficient to satisfy him that he was a man of distinguished
appearance, tall and well formed though slight, and clothed as one
of the higher classes.

The ascent was somewhat laborious but not long, after they had
once discovered the right road; and about twenty minutes brought
the party to an old bridge and gate under a deep arch. By the faint
light of the candle, which was by this time wellnigh burned out, the
place looked fully as ruinous and desolate as the peasants had
represented it to be. The rugged outlines of some of the towers
showed that much of the masonry had fallen, and the key-stone of
the arch and a large mass of rubbish only left room for the horses to
pass one at a time. Still, however, the light they had seen from
below continued to stream from three windows in a great, dark,
shapeless mass of buildings, and the approach of the new-comers
did not seem to have been discovered by the persons within, if there
were any.
Stop a moment, said Edward, pausing under the arch. As we do
not know what sort of persons we shall find within, it is well to be
prepared. The priming of my pistols may be damp, though the
holsters are made as tight as possible. And, standing under the
shelter of the walls, he took the weapon from his saddle-bow, threw
the powder out of the pans, and primed them anew. He then took
the very useful precautions of ascertaining that no water had
entered the barrels and that the balls were still in their places.
Ay, he has got two lives there, said Pierrot, keeping close to his
master; and then, fastening the horses to some chains which hung
about the bridge, the whole party advanced toward the building in
which the lights were seen. A low and narrow door admitted them to
the foot of a small stone stair-case, and, lighted by the blinking
lantern, they began to ascend. They had hardly gone half-way up—
Edward with one pistol in his belt and the other in his hand—when
they heard a clear, merry peal of laughter; and, somewhat hurrying
his pace, lest the little candle should go out before they reached the
object of their search, the young Englishman reached a little ante-
room with a door on the opposite side, through the large key-hole of
which a ray of light streamed out upon the floor.

The door was thrown open without ceremony; but the scene which
the interior of the large hall or chamber presented was what none of
the party expected. Seated round a table, on which were the
remains of an abundant meal, with plenty of wine, and sundry
papers and maps, was a party of gentlemen, richly dressed, with the
exception of one who occupied the top of the board and who was
habited as an ecclesiastic. A gentleman on the abbé's right hand was
in the very act of speaking with some gesticulation when the door
was flung open; but he instantly stopped. The party at the door
stopped, also, in much surprise, and each group gazed upon the
other for a moment in silence.

CHAPTER XXVII.
The hall was lighted by three large sconces hung against that part of
the wall nearest to the table; but still the extent of the chamber
rendered the light feeble, except immediately under the burners. It
cannot be said that the appearance of Edward Langdale and his
companions was very prepossessing. Edward himself wore his hat
and plume, which had been thrown off before he plunged into the
water; but his dress was soiled as well as wet. The stranger whom
he had saved was in a still worse plight: his hat, of course, had been
lost in his struggle with the torrent, and his forehead and part of his
face were covered with dripping locks of long black hair. His sword,
which had remained in the sheath, was the only distinguishing mark
of a gentleman about him. Pierrot and Jacques Beaupré looked far
more like bravos than the followers of an English gentleman of those
days; and the two ill-favored blacksmiths, one armed with a half-
extinguished lantern and the other with a sledge-hammer, did not
add to the beauty or respectability of the group.
No wonder, then, that several of the gentlemen at the table laid their
hands upon their swords; and the one who had been speaking
advanced a step or two, exclaiming, in a threatening tone, What is
this? What means this ill-mannered intrusion? Who are you, sirs, and
what seek you here?
Shelter from the storm, and food, if it can be procured, said
Edward: we know not upon whom——
But, before he could finish the sentence, the gentleman to whom it
was addressed started forward and caught him by the hand,
exclaiming, What! Ned, my boy! How came you to seek me here?

I did not seek you here, my lord, replied Edward, and, to say
truth, if I had known you were here, I should not have come. I was
on my way to Aix to join your lordship, according to your commands;
but the road is impassable. Some of us have been half drowned;
and, though this is a desolate-looking place, we said, 'Any port in a
storm.'
But who are these gentlemen with you? asked Lord Montagu, still
speaking in French, but running his eye somewhat doubtfully over
the group of five persons who had advanced some way from the
door.
Those two, answered Edward, in the same gay tone, which was
generally affected by pages of noble houses,—those two are my
servants, or rather your lordship's, the renowned and reformed
Pierrot la Grange and the facetious Jacques Beaupré. Those two—
the one with the lantern and the other with the hammer—are two
respectable blacksmiths and horse-doctors, who have joined
themselves on to me and mine and did good service in curing one of
my horses. They profess to be Savoyards returning to their own
country.
They shall be welcome, said Lord Montagu, smiling,—most
welcome, for I have no less than five good horses sick of some
distemper at Chambéry. But who is the other,—that gentleman who
seems half drowned?
He was half drowned a few minutes ago, my lord, replied the
youth, and so was I; but he will probably tell you more of himself if
you will ask him. His horse leaped with him into the river, and it was
a hard matter to get him out.
I hold it but courteous in these bad times, said Lord Montagu, to
follow the old knightly rule and ask no stranger any questions,—
before he has cut your throat; and therefore we will invite him to
sup, and leave him to explain himself. He seems a gentleman.

Yes, my lord, was all Edward's reply; but a very peculiar expression
crossed his countenance as he uttered those three words, which,
had Lord Montagu seen it, might have caused more inquiry. That
nobleman, however, had turned to speak for a moment with the
gentlemen who had been seated with him; and he then advanced to
the stranger, inviting him courteously to be seated and take some
refreshment, and expressing sorrow for the accident which had
befallen him. He also bade the other four sit down and eat; and,
there being no place for so many at the table, filled as it was, most
of those who had already supped rose and gathered together at the
end of the board, Edward taking his place amongst them without
touching any thing.
Lord Montagu introduced him to the rest in kind terms, saying, My
page and young friend, Monsieur Edward Langdale, Monsieur le
Prince de ——, Monsieur le Comte de ——, Monsieur l'Abbé Scaglia,
the Duke of Savoy's prime minister. We came here on a little party of
pleasure, Ned, and sat long over our cups, in truth, hardly hearing
that the storm was still going on. Come, my good youth, sit down
and eat. You must be well weary of all the adventures which the fair
duchess writes me you have gone through. Eat, boy! eat!
Your pardon, my lord, said Edward, gravely: I will take a cup of
wine here standing: that is all. I have much to tell your lordship.
By-and-by, by-and-by, said Lord Montagu, we shall have plenty of
time and plenty to talk of. Well, drink if you will not eat.
Edward Langdale advanced to the table, filled himself a goblet of
wine, and returned with it to Lord Montagu's side. Before he could
raise it to his lips, however, the stranger whom he had saved from
drowning turned round his head, saying, with a polite smile, Let me
have the pleasure of drinking with you, young gentleman, in
memory of the service you rendered me. I do not know your name,
though your face is very familiar to me.

A dark cloud gathered upon Edward Langdale's brow, and he
answered, not sharply, but with stern, cold bitterness, I neither eat
with you nor drink with you, sir.
The stranger started up with his face all on fire, and exclaimed, with
his hand upon the hilt of his sword, Do you mean to insult me, sir?
I mean to tell you, sir, said the youth, boldly, that I am Edward
Langdale,—your father's son; and that you have robbed me of that
to which neither he nor you had any right,—my sweet mother's
estates.
Robbed? robbed? cried Sir Richard Langdale, furiously drawing his
sword.
Ay, robbed,—swindled, if you like it better, said Edward. Put up
your sword, or sheathe it here, he continued, throwing his arms
wide open and exposing his chest. I do not fight with my brother.
The other rushed upon him like a madman.
What is this? what is this? cried the Abbé Scaglia, running forward.
Back, madman! exclaimed Lord Montagu, seizing Richard Langdale
by the collar.
Pierrot la Grange also darted forward and tried to push between. But
all were too late. Edward fell to the ground with a heavy fall, and his
brother withdrew his sword all dripping with blood.
The burly blacksmith advanced toward him with his hammer raised
in the act to strike him on the head, exclaiming, in very good French,
The murdering villain! He has killed the man who saved his life at
the risk of his own, not an hour ago!
But Lord Montagu caught his arm, saying, Stand back. This must be
inquired into by justice. No more slaughter here. Sir, give up your
sword! You are a prisoner.

Aid, all men, to arrest him! cried the Abbé Scaglia. I command
you in the duke's name!
Sir Richard Langdale moved not a muscle, but stood gazing at the
fallen form of his brother with a face as pale as marble and bloodless
lips. Such sudden changes of feeling will often take place in terrible
circumstances. When the dreadful deed, prompted by the fierce fire
of passion, is once done, we know all its horrors; but not before.
The consummation is like the lightning-flash upon a corpse, showing
every ghastly feature more livid and frightful from the remorse-like
glare that darts across it. Suddenly he started, raised his hands to
his head, tearing his long black hair, and exclaiming, Curse the
lands! Curse the riches!
Here! cried Lord Montagu, take him away, you two. Guard him
safely, but do him no hurt. You stout fellow, aid us to raise this poor
lad, and let us see if nothing can be done for him. On my life, I
would as soon have lost my brother!
Let me tend him, sir, said the blacksmith with the lantern: I have
cured many a horse as bad hurt as he; and a horse and a man are
much the same thing.
Not quite, said Lord Montagu, who even at that moment could not
altogether resist the joking spirit of his times and his party. Heaven!
how he bleeds! Gentlemen, he was the noblest lad—the promptest
with hand and head and heart—I ever saw. Poor Edward! can we do
nothing for you?
As he spoke, they raised the youth and laid him on the table, and
the blacksmith tore open his vest. The movement seemed to awaken
him a little; and, probably with thoughts far distant, he exclaimed, in
a faint voice, No, never! no, not with life! But the rough hands
stayed not their work; and, after gazing for an instant at his
wounded side, the man turned to his companion, saying, Ivan, run
down and bring up the pack, quick! We can stop this bleeding. Do

you not see? it does not jerk. Then, if none of the vitals be touched
——
A hundred crowns if you save him till we can get to Aix, said Lord
Montagu.
I think I can save him altogether, said the man. The thing is,
people will not treat man as if he were a beast; and so they kill him.
Man and beast are only flesh, and all flesh is grass.
But it is needless to discuss or to display any further the views and
principles of Edward's somewhat rough doctor, or to detail the
treatment he underwent. There was the usual amount of bustle and
confusion, and the much talking and the recommendation of many
remedies which could not be procured and would have done no
good if they had been there. Suffice it that the bleeding was soon
stayed, and that Edward recovered from the fainting-fit into which
the wound, probably penetrating some very sensitive part, had
thrown him. The blacksmith by no means wanted mother-wit, and
his treatment was probably based upon the sound principle of
merely aiding nature. The lad spoke a few words, and they tried to
impose silence upon him; but he would not hold his peace till those
around assured him that no one had hurt his brother and that he
was safe in another chamber.
All Lord Montagu's anxiety seemed to be to get him to Aix; and he
went out himself and sent out more than once to see if the storm
was over. Luckily for Edward, it continued all night and part of the
next morning; I say luckily, for the hands in which he was were
probably better calculated to bring about his recovery than any
which could have been found in a small town in Savoy, as medical
science went in those times.
In the mean while, the party assembled made themselves as
comfortable as they could in disagreeable circumstances of many
kinds; and the heavy tread of Sir Richard Langdale was heard
through the night beating incessantly the floor of the room above.

Toward morning that wearisome footfall ceased, and Lord Montagu,
who sat by Edward's side and was still awake, said to himself, That
poor wretch has found sleep at length. Now, which is the happiest?
—he, or poor Ned here? I would rather be that boy than the man
who has killed his own brother. They say that Edward saved his life,
too, not an hour before. Very likely! He is fit for any gallant act.
Heaven! what must that man's thoughts be?
Soon after, the Abbé Scaglia roused himself in the corner where he
had ensconced him, and, moving quietly up, talked in a low tone for
some twenty minutes with Lord Montagu. They then roused the rest
of the party who had been supping there, and went down into the
court-yard, where they found the horses of Edward Langdale and his
companions. Their own were hidden in one of those deep vaults
under the great tower which were common in most feudal castles,
especially in border-districts, as a safe and silent receptacle of stolen
cattle and horses.
Though it was still raining, most of the party mounted and rode
away, promising to send up a litter and a surgeon as soon as the
road was passable. Lord Montagu himself said he would remain with
the poor lad, and reascended to the chamber where he had left him.
All was silent there: the wounded youth had fallen into a sleep which
seemed calm, and the two blacksmiths were nodding beside him.
The English nobleman then went up to the floor above, where he
found Jacques Beaupré asleep across the door, and Pierrot sitting
up, but rubbing his eyes as if he had not been long awake.
In answer to the nobleman's questions, Pierrot detailed all that had
occurred upon the road, and dwelt upon the gallant conduct of his
young master. He little thought, said the man, that he was risking
his own life to save the very man who would kill him. But I have
often heard say that it is unlucky to rescue a man from drowning. As
to this man in here, sir, I believe he is mad; for he has been walking
about all night,—sometimes talking to himself, sometimes groaning
as if his heart would break. I had better wake him, perhaps.

No, no! Let him sleep if he can, said Lord Montagu, quickly. Well
may he groan! Pray Heaven neither of us may ever have such cause,
my man. When you hear him move, get him some wine. There is still
some down-stairs. Till then, let him alone. If he sleeps, it is the best
thing for him.
Thus saying, he went down again, and, finding every thing as
before, approached the window and gazed at the morning light, still
pale and blue, spreading up from the mountain-edges into the rainy
sky. After about half an hour, Edward turned painfully and asked for
some water. His lord gave it to him with a kindly word or two, and
the blacksmiths woke up and examined the wound. They seemed
satisfied with its appearance, and one of them said, loud enough for
Edward to hear, He will get well, sir.
Oh, what a blessed thing is hope! Those few words were a better
balm than any druggist could have supplied. They brought with
them, too, the thought of Lucette; and, beckoning to Lord Montagu
to hold down his head, he whispered, If I should die, my lord, I
beseech you to write a few lines to the old Marquise de Lagny, to tell
her the fact. She will be with the court of France, wherever that may
be.
No, no; you will get well, Ned, said Lord Montagu, in a cheerful
tone. I do not intend to part with you yet. But now you must
positively be silent if you would not increase the evil.
Some four or five hours passed. The rain cleared away, the sun
broke out, and Lord Montagu looked anxiously from the windows
which were turned toward the road, in expectation of the promised
litter. All he could see, however, was a large party of Savoyard
peasantry working hard, apparently, to remove some obstruction
from the highway.
He was still gazing forth, when Pierrot appeared at the door, and,
finding all still, beckoned to him.

My lord, he said, in a low voice, when Montagu had joined him, I
can hear nothing of that man above, nor Jacques either. He could
not get out of the windows; and I should not wonder if he has
hanged himself.
Lord Montagu started and instantly ran up-stairs, thinking the
conclusion at which Pierrot had jumped not at all improbable. He
opened the door gently and looked around. The sun was shining full
into the room, but Sir Richard Langdale was not there. The only
thing that could indicate the mode of his escape was a pair of large
riding-boots, very wet, which lay on the floor; and it is probable that,
opening the door cautiously while the two men were asleep, he had
stepped lightly over them and then gone down the stairs.
What a thing is the love of life! thought Lord Montagu. This man
would rather live miserable than risk the grave. However, I cannot be
sorry; and I believe poor Ned will be glad.
He entered the room below as silently as possible; but Edward, who
had heard his rapid step running up the stairs, turned his head,
asking, Is there any thing the matter above?
Only that your brother has escaped, said his lord.
Thank God! said the young man, with a smile. Pray, do not pursue
him, my lord.
I will not, replied Montagu: make your mind easy, Ned.
Here come some people with a litter up the hill, said one of the
blacksmiths.

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Digital Television Mpeg1 Mpeg2 And Principles Of The Dvb System 2nd Edition Herv Benot

More Related Content

Similar to Digital Television Mpeg1 Mpeg2 And Principles Of The Dvb System 2nd Edition Herv Benot

Recently uploaded

Digital Television Mpeg1 Mpeg2 And Principles Of The Dvb System 2nd Edition Herv Benot