dr. nikos desypris [email protected] [email protected] feb. 2011 postgraduate course...

Dr. Nikos Desypris

[email protected]

[email protected]

Feb. 2011

Postgraduate course

University of Athens

5th Lecture on Digital TV

Academic Writing (separate presentation)

DTV Transmission and Reception System

DTV Transmission Systems

DTV Transmission & Reception System

Layered approach to system design Broadcasting system: includes four subsystems:

Presentation / Compression / Transport – Multiplex / Transmission

Distribution channels:

Traditional: Terrestrial / Cable / Sattellite

New: Internet TV / IPTV / Hand-held devices

DTV receiver: the broadcast sequence is reversed

Reception / Transport – Demultiplex / Decoding / Presentation

Layered approach to system design

Broadcasting system: 4 sub-systems

Presentation: image format and sound

Compression: reduces data rate to practical levels

Transport multiplex: packetizes audio, video and data: includes assembly instruction for the DTV receiver

Transmission: adds error correction and modulates symbols for

channel distribution

The DTV Receiver: 4 sub-systems The sequence is reversed Reception:

Signal capture, demodulation and recovery of data packets from the transmission channel

Transport demultiplex: extraction of audio, video and data packets and assembly instructions

Decoding: Expansion of compressed audio and video data

Presentation: audio and video are synchronised and the complete program is ready for viewing

Layered approach to system design

Engineering Basics

Voltage levels

Noise

Units of data Bits, Bytes and Words

Repetive signals: frequency and the Hertz

Analogue to Digital Conversion and Quantisation

Data Structure and Processing Packets, Headers and Payloads; Streams; Data Buffer and

Parsing

Engineering Basics Voltage levels

For digital systems, discrete voltage level represent information. Two levels are used for 0 and 1; however 3, 4 or more distinct levels can be used

NoiseCan be internal or caused from external sources. Becomes a problem if amplified

Units of databit: 0 or 1byte: 8bits, 256 levelsword: two bytes, 16 bits, 65536 valuesdouble word: four bytes, 32 bits, 232 valuesin digital audio applications: 18 and 20 bits wordsin digital video applications: 10 and 12 bit words are common

Engineering Basics

Repetition signals: the notion of the frequencyUnit: one cycle per sec referred as Hertz named after the German physicist Heinrich Rudolf Hertz (1857-1894)

Definition of 1Hz, 1KHz, 1MHz, 1 GHz

Audio signals are between 20 Hz and 20KHz.

Video images have spatial (within a frame) and temporal (between successive frames) frequency characteristics

Engineering BasicsAnalogue to Digital Conversion and Quantisation

Engineering BasicsData Structure and Processing Bits and bytes are arranged in groups referred as data structures or object model, which are processed by machinesPackets, Headers and PayloadsData is grouped into packets having the actual content of transmission (the payload) and additional header information describing the payloadStreamsContinuous transmission of data (data stream): can be asynchronous (at random), synchronous (with a defined timing) or isochronous (defined amount of data over a time period)Data BufferA storage device, often RAM that stores and transmits data at different rates.ParsingExamining a data stream for defined values in a data structure and extracting desired information is called parsing


1. Presentation: image format and sound

2. Compression: reduces data rate to practical levels

3. Transport multiplex: packetizes audio, video and data: includes assembly instruction for the DTV receiver

4. Transmission: adds error correction and modulates symbols for


Broadcasting System 1: Presentation General Visual Perception:

Luminance sensitivity, Colour Sensitivity

Video Presentation formats: Aspect Ratio Scanning methods Refresh Rate Resolution and pixel grid Synchronisation Pixel Squareness Sampling and colorimetry

Audio Formats: Audio Perception Audio Presentation Formats

Broadcasting System 1: Presentation

General

Visual Perception

Video Presentation formats

Audio Formats


General

Presentation is about Immersive consumption; HDTV was envisaged as a large-screen theatrical

viewing experience Importance of surround sound in the immersive

experience

DTV is a mass product at an affordable price

based on current technological limits


General

Visual Perception


Audio Formats


Visual Perception

General

Luminance sensitivity,

Colour Sensitivity

Broadcasting System 1: Presentation Visual Perception

General


Visual Perception Luminance sensitivity

Y= 0,30 * Red + 0,59 * Green + 0,11 * Blue

Colour Sensitivity Combinatorial process of mixing red, green and blueVisible light is between 380 nm and 760 nmRed -> 440 nmGreen -> 540 nmBlue -> 700 nm


General

Visual Perception


Audio Formats


Video Presentation formats:

General

Aspect Ratio

Scanning methods

Refresh Rate

Resolution and pixel grid

Synchronisation

Pixel Squareness

Sampling and colorimetry


Video Presentation formats: General

A video format consists of :

- Aspect ratio

- Raster (pixel grid structure)

- Scanning method

- Refresh rateNote: Analogue TV had only one video format and one

colour space sampling structure. There are 18 different DTV formats in the ATSC standard (many more in satellite and cable!)

Broadcasting System 1: Presentation Video Presentation formats: Aspect Ratio

Broadcasting System 1: Presentation Video Presentation formats: Scanning methods


Video Presentation formats: Refresh Rate

Movie industry 24 frames/sec

TV: 30 Hz (half the AC 60 Hz power line frequency)

Other frame rates: 29.97, 30, 60, 59.94

Film rates: 24 or 23.98


Video Presentation formats: Resolution and pixel grid

Always defined as pixels per line by pixels per frame

Refresh rate defines the number of complete frames;

60p: 60 progressive frames per second

60i: denotes 30 frames per second, each frame comprising

of two interlaced fields (one field odd lines, the other even

lines)Important note: The no of pixels and lines in an image format is

not a statement of the image resolution which is dependent on the entire signal processing chain

Broadcasting System 1: Presentation Video Presentation formats: Synchronisation

The timing relationship of the scanning process that produces pixels, lines and a display raster (pixel grid) must be unambiguously established.

1920 pixels x 1080 lines 1280 pixels x 720 lines

Define the active pixels and lines in a display, while the actual scanning number for each video presentation format is

2200 pixels x 1125 lines and1650 pixels x 750 lines respectively

Broadcasting System 1: Presentation Video Presentation formats: Synchronisation

DTV eliminates blanking intervals during compression


Video Presentation formats: Pixel SquarenessSquare pixels are defined (in DTV standards) as when the

number of horizontal to vertical pixels is the same proportion as the display aspect ratio.

640:480 = 4:3, HD formats 1920x1080 and 1280x720 are 16:9, therefore HD formats employ square pixels.

However SD, 720x480 and 704x480 are not 4:3 and it is quite important for creating graphics on video.

A simple solution is to create a graphics canvas that is the same display format as the intended display

Broadcasting System 1: Presentation Video Presentation formats: Sampling

Broadcasting System 1: Presentation Video Presentation formats: Colorimetry


General

Visual Perception


Audio Formats


Audio Formats:

Audio Perception

Audio Presentation Formats

Broadcasting System 1: Presentation Audio Formats: Audio Perception

Audible sound for young people: 20Hz to 20KHz

Broadcasting System 1: Presentation Audio Formats: Audio Presentation Formats

Broadcasting System 2: Compression

Compression MPEG-2 Video Compression

Exploiting Redundant Visual Information Group of Pictures Compression Steps Exploiting Temporal Redundancy

Audio Compression Aural Perception and Compression Algorithms

Masking Framing


Compression

MPEG-2 Video Compression

Audio Compression

Broadcasting System 2: CompressionFull bandwidth HDTV picture with 6 audio channels is 1,5 GBps.

To fit in a 6MHz channel, data reduction of 50:1 for video and 12:1 for audio are common.

Compression generation: an encoding and decoding cycle.

Each compression generation creates artefacts

Consumer electronics are not to highest fidelity; however they have consumer satisfaction

Compression ‘engines’ are divided in two broad categories:

Lossless compression reduces the volume of data and, when reconstructed restores it to its original state, perfectly without any loss of information

Lossy compression discards data based on auditory and visual sensory characteristics and limits

Broadcasting System 2: CompressionRecent new video codecs:

VC1 derived from Windows Media and MPEG-4 Part 10 claim to deliver the same quality of MPEG-2 with half the rate

MP3 is part of MPEG-1 or MPEG-2 Audio Layer III

Advanced Audio Coding (AAC) is also an MPEG standard

AC-3 (Audio Codec 3) with support of up to 6 channels was intended for theatrical applications

Key dataJPEG focuses on still imagesMPEG (Motion Pictures Expert Group)MPEG1: 1988-1991MPEG2: 1994MPEG4, Part 10: latest version of mpeg


Compression


Audio Compression



General

Exploiting Redundant Visual Information

Group of Pictures

Compression Steps

Exploiting Temporal Redundancy

Broadcasting System 2: CompressionMPEG-2 Video Compression: General

Compression is characterised by a ratio (video 50:1, audio 12:1)

The higher the ratio the fewer the number of bits

Raw bit rates for HD video are about 1,5 Gbps. With 4:2:2 sampling and 50:1 compression, yields 30 Mbps. With 4:2:0 sampling (1 Gbps) and 50:1 compression results 20 Mbps.

The higher rate (30Mbps) can survive a generation or two of editing without artefact, while the lower rate (20Mbps) is used for presentation only.

For multiple generations of compression and decompression in production environment, video should be at least 100 Mbps for HD and 40 Mbps for SD.

Broadcasting System 2: CompressionMPEG-2 : Exploiting Redundant Visual Information

Coding only scene differences results in significant compression, but depends on scene complexity, for example:

High: Detail -> spatial complexity: rain grounds

Motion: Change -> temporal complexity: sports

Low: Still = simple: talking head, stationary text and graphics

MPEG-2: two attributes for scene complexity: profile and level:Profile compression complexity related to colour sampling:Simple; Main; 4:2:2 , SNR, Spatial and HighLevel specifies picture size in pixels:Low; Main; High 1440 and High

Broadcasting System 2: CompressionMPEG-2 : Exploiting Redundant Visual Information

SD is MP @ML

HD is MP @HL

Profile

Level

Broadcasting System 2: CompressionMPEG-2 Video Compression: Group of Pictures I frames: Intraframe, encoded independent of other

frames P frames: Predictive, encoding is dependent on B frames: Bidirectional, encoding is dependent on

previous or subsequent I or P frames A ‘long GOP’ is usually up to 15 frames such as:

I B B P B B P B B P B B P B B,corresponding to 0,5 sec on 1080i at 30Hz, or 0,25 sec for 720p at 60Hz (important for switching TV channels)

I frames are called anchor frames


MPEG-2 Video Compression: Compression StepsCompression of an I-frame: It is divided into 8x8 pixels blocks Discrete cosine transform (from 8 bits pixels to 11 bits

coefficients!) Weighting and Requantisation (to reduce high frequency

coeffs) Variable length coding (to assign small codewords to most

common symbols as in Morse code where E, T -> . and _) Run length coding (to code a sequence of same digits more

efficiently)


MPEG-2 Video Compression: Exploiting Temporal Redundancy

I frame – current frame = P Frame (different pixels not values!)

frame 2 – I frame – P frame = B frame


P: difference 1 and 3B: difference 2, I and P>

P > B


Compression


Audio Compression


Audio Compression Aural Perception and Compression Algorithms, based on

psychoacoustic model• Masking:

When one sound hides another sound, it need not be encoded

• Framing:Further data reduction is taking place by requantising high

frequencies using the psychoacoustic model


Presentation: image format and sound

Compression: reduces data rate to practical levels

Transport multiplex: packetizes audio, video and data: includes assembly instruction for the DTV receiver

Transmission: adds error correction and modulates symbols for


Broadcasting System 3: Transport Multiplex

Compressed Data Streams

Packetised Transport Multiplex Concepts PSI and Program Assembly

Transport Stream Multiplex Statistical Multiplexing and Rate Shaping



Packetised Transport


Compressed Data StreamsThe result of compression process is a continuous flow of audio

and video data known as an Elementary Stream (ES)

ES become part of Packetized Elementary Streams for not loosing audio/video timing relationship



Packetised Transport


Packetized Transport

Two types of MPEG streams: Program and Transport

Program streams:are used in environments that rarely corrupt data (such as a DVD player) andcontain a single audio and video program

Transport streams:Enable robust delivery of data over noisy channels and error prone environmentsmay contain numerous programspackets are much smaller that those in program streams (only 188 bytes)


Packetized Transport: Multiplex concepts


Packetized Transport: PSI and Program Assembly


Packetized Transport: Transport Stream Mulitplex


Packetized Transport: Statistical Multiplexing & Rate Shaping

Broadcasting System 4: Transmission

Data Protection Randomisation Reed Solomon Encoding Data Interleaving Trellis Encoding

Digital Modulation Modulation

Digital Modulation Analogue Modulation Digital Modulation Vestigial SideBand Modulation (VSB) Quadrature Amplitude Modulation (QAM) Quartenary Phase Shift Keying (QPSK)

Broadcasting System 4: Transmission Data Protection

Randomisation: Spreads data over the full transmission channel Scrambling the stream avoiding repetitive patterns

Reed Solomon Encoding Use of forward error Correction Code for correcting

multiple bit errors, since retransmission is not possible Data Interleaving

Spreading data over time in order to avoid lost sequences due to impulse noise (i.e. lightning)

Trellis Encoding Transforming bits to symbols used to modulate the RF

carrier (generally depends on the modulation standard)

Broadcasting System 4: Transmission

Broadcasting System : Summary

Four sub-systems form a DTV broadcasting system: Presentation, compression, transport and transmission

Exploitation of the characteristics and limitations of the sensory perception are used in DTV presentation and compression systems

Packetized audio and video, along with assembly instructions are sent to a DTV receiver

Digital transmission uses sophisticated method of error protection and correction

dr. nikos desypris [email protected] [email protected] feb. 2011 postgraduate course...

Documents