L64005 MPEG-2 Audio/Video Decoder Technical Manual
1-7
Final Rev F
Copyright 1996 by LSI Logic Corporation. All rights reserved.
recover by waiting for the next slice, without having to drop an entire
picture.
Slices consist of a header and one or more
macroblocks. The slice
header contains position and quantizer scale information.
A macroblock is the basic unit for motion compensation and quantizer
scale changes. In MPEG-2 the block can be either eld or frame coded.
Each macroblock consists of a header and six component 8 x 8 blocks:
four blocks of luminance, one block of Cb chrominance, and one block
of Cr chrominance. The macroblock header contains quantizer scale and
motion compensation information.
A macroblock contains a 16-pixel by 16-line section of luminance com-
ponent and the spatially corresponding 8-pixel by 8-line section of each
chrominance component. A skipped macroblock is one for which no DCT
information is encoded.
Blocks are the basic coding unit, and the DCT is applied at this block
level. Each block contains 64 component pixels arranged in an 8 x 8
order. Note that pixel values are not individually coded, but are compo-
nents of the coded block.
Note that the picture area covered by the four blocks of luminance is the
same as the area covered by each of the chrominance blocks. Each lumi-
nance pixel corresponds to one picture pixel, but because the chromi-
nance information is subsampled with a 2:1 ratio both horizontally and
vertically (4:1 total), each chrominance pixel corresponds to four picture
pixels.
1.1.3
Video Decoding
Video decoding is the reverse of video encoding and is intended to
reconstruct a moving picture sequence from a compressed, encoded bit-
stream. Decoding is simpler than encoding because there is no motion
estimation performed and there are far fewer options.
The data in the bitstream is decoded according to the syntax dened in
the MPEG-2 standard. The decoder must rst identify the beginning of a
coded picture and identify the type of picture, then decode each individ-
ual macroblock within a particular picture. Motion vectors and macroblock
types (each of the picture types I, P, and B have their own macroblock
types) present in the bitstream, are used to construct a prediction of the