PAL uncompressed• 768x576 pixels per frame

x 3 bytes per pixel (24 bit colour) x 25 frames per second

• ≈ 31 MB per second• ≈ 1.85 GB per minute

191–192

NTSC uncompressed• 640x480 pixels per frame

x 3 bytes per pixel (24 bit colour) x 30 frames per second (approx)

• ≈ 26 MB per second• ≈ 1.6 GB per minute

191–192

In the camera – DV + Firewire In the computer – video capture card Digitization in camera (DV) means

less noise Less noise allows better compression

193–195

Play back a video stream as it arrives over a network (like broadcast TV), instead of downloading an entire video clip and playing it from disk (like renting a DVD)

197

Start playing a downloaded clip as soon as enough of it has arrived

Starts when the (estimated) time to download the rest is equal to the duration of the clip

198

Digital video devices must conform to standards

Digital standards must maintain compatibility with older analogue standards for broadcast TV

199

Required for TV, so encountered in captured footage

Each frame is divided into two fields Field 1: odd lines; Field 2: even lines Fields are transmitted one after the

other Frame is built out of the interlaced

fields

200

PAL (Phase Alternating Line)• Western Europe, Australia & New Zealand,

China,… NTSC (National Television Standards

Committee)• North America, Japan, Taiwan, parts of

South America,…

199

SECAM (Séquential Couleur avec Mémoire)• France and former Soviet Union• Standard only used for transmission• Uses PAL cameras etc

199

PAL• Frame has 625 lines, 576 are picture• 25 frames (50 fields) per second

NTSC• Frame has 525 lines, 480 are picture• 29.97 frames (59.94 fields) per second

(Often quoted as 30 frames per second)

202

Digital video standard, properly called Rec. ITU-R BT.601

720 luminance samples (Y), 2x360 colour difference samples (B−Y and R−Y) per line

PAL 720x576 pixels; NTSC 720x480 pixels

Pixels are not square

202–203

Used in CCIR 601 Twice as many Y

samples as each of the colour difference samples

Co-sited: same pixel is used for all three samples

Reduces data rate to just over 20MB per second

203

Consumer format, also known as mini-DV

DVCAM, DVPRO use different tape formats, but generate the same data stream

4:1:1 chrominance sub-sampling Data rate constant 25Mbits per

second Compression ratio 5:1

204

PAL DV 4:2:0 chrominance sub-sampling

210

NTSC DV 4:1:1 chrominance sub-sampling

210

ISO/IEC Motion Picture Experts Group Series of standards including

• MPEG-1 intended for video CD• MPEG-2 used in DVD and broadcast• MPEG-4 for low bitrate multimedia

204–206

Profiles define subsets of the features of the data stream

Levels define parameters such as frame size and data rate

Each profile may be implemented at one or more levels

Notation: profile@level, e.g. MP@ML

204–205

MPEG-2 Main Profile at Main Level (MP@ML) used for DVD video• CCIR 601 scanning• 4:2:0 chrominance sub-sampling• 15 Mbits per second• Most elaborate representation of MPEG-2

compressed data

205

Designed to support a range of multimedia data at bit rates from 10kbps to >1.8Mbps

Applications from mobile phones to HDTV

Video codec becoming popular for Internet use, is incorporated in QuickTime, RealMedia and DivX

205–206

Visual Simple Profile (SP), suitable for low bandwidth streaming over Internet

Visual Advanced Simple Profile (ASP) suitable for broadband streaming

SP@L1 (Level 1 of Simple Profile), 64 kbps, 176x144 pixel frame

ASP@L5, 8000 kbps, full CCIR 601 frame

205

Spatial (intra-frame) compression• Compress each frame in isolation,

treating it as a bitmapped image Temporal (inter-frame)

compression• Compress sequences of frames by only

storing differences between them Always some compression

because of sub-sampling

206–208

Image compression applied to each frame

Can therefore be lossless or lossy, but lossless rarely produces sufficiently high compression ratios for volume of data

Lossless compression implies a loss of quality if decompressed then recompressed

Ideally, work with uncompressed video during post-production

207

Key frames are spatially compressed only• Key frames often regularly spaced (e.g. every

12 frames) Difference frames only store the

differences between the frame and the preceding frame or most recent key frame

Difference frames can be efficiently spatially compressed

207–208

Purely spatial compression Apply JPEG to each frame Used by most analogue capture

cards No standard, but MJPEG-A format

widely supported

209–210

Starts with chrominance sub-sampling of CCIR 601 frame

Constant data rate 25Mbits per second Higher quality than MJPEG at same rate Apply DCT, quantization, run-length

and Huffman coding on zig-zag sequence – like JPEG – to 8x8 blocks of pixels

210–211

If little or no difference between fields (almost static frame), apply DCT to block containing alternate lines from odd and even fields

If motion between fields, apply DCT to two 8x4 blocks (one from each field) separately, leading to more efficient compression of frames with motion

210–211

Shuffling• Construct video segments by taking 8x8

blocks from five different areas of the frame, to ‘average’ amount of detail

• Calculate coefficients for whole video segment, making more efficient use of available bytes

210–211

Spatial compression based on quantization and coding of DCT coefficients

Temporal compression based on motion compensation• Record displacement of object plus changed

pixels in area exposed by its movement

211–212

I-pictures purely intra-frame compressed

P-pictures ‘predictive’• Difference frames based on earlier I- or P-

pictures B-pictures ‘bi-directionally predictive’

• Difference frames based on preceding and following I- or P-pictures

212–213

Group of Pictures (GOP)• Repeating sequence of I-, P- and B-pictures• Always begins with an I-picture• Display order – frames in order they will be

displayed• Bitstream order – re-ordered so that every

P- or B-picture comes after frames it depends on, allowing reconstruction of the complete frames

213–214

Source Input Format (SIF)• 4:2:0 chroma sub-sampled• 352x240 pixel frame

MPEG-1 compressed SIF video at 30 frames per second has data rate of 1.86Mbits per second (CD video – 40mins of video at that rate)

MPEG-1 can be scaled up to larger frames, but cannot handle interlacing

214

Standard defines an encoding for multimedia streams made up of different sorts of object – video, still images, animation, 3-D models…

Higher profiles divide a scene into arbitrarily shaped video objects each one may be compressed and transmitted separately, scene is composed at the receiving end by combining them

SP and ASP restricted to rectangular objects, usually complete frames

215

Refinement of MPEG-1 compression• I-pictures compressed by quantizing and

Huffman coding DCT coefficients• Improved motion compensation leads to

better quality than MPEG-1 at same bit rates

215

Simple Profile• P-pictures only• Efficient decompression, suitable for PDAs

etc Advanced Simple Profile (ASP)

• B-pictures• Global Motion Compensation• Sub-pixel motion compensation

215–216

Cinepak – Longest established, high compression ratio, takes much longer to compress than to decompress

Intel Indeo – Similar to Cinepak, but roughly 30% faster compression

Sorenson – More recent, higher quality and better compression ratios than other two

All three based on vector quantization Quality of all three inferior to MPEG-4

216–219

Divide each frame into small rectangular blocks (’vectors’)

Code Book – collection of constant vectors representing typical patterns (edges, textures, flat colour,…)

Compress by replacing each vector in image by index of vector from code book that most closely resembles it

216

Making a constructed whole from a collection of parts• Selection, trimming and organization of raw

footage• Apply transitions (e.g. dissolves) between

shots• Combination of picture with sound

No changes made to the footage itself

223–230

Changing or adding to the material• Most changes are generalizations of image

manipulation operations (e.g. colour correction, blurring and sharpening,…)

Compositing – combining elements from different shots into a composite sequence

Animating elements and combining animation with live action

230–236

Compromises required to bring resource requirements of video within capabilities of delivery media (e.g. networks) and low-end machines• Reduce frame size (e.g. downsample to

quarter frame)• Reduce frame rate (12fps is OK for smooth

motion, flicker not a problem on computer)• Reduce colour depth

236–237