Lecture 14
MPEG-2
Enhancements over MPEG-1
MPEG-2 has several enhancements over MPEG-1, and it is designed for broadcast use as well as computer use. Some of the enhancements include:
Encoding Fields in MPEG-2
Remember, TV signals are interlaced. That means that the TV first reads and displays all the odd scan lines before reading and displaying the even scan lines for a frame. MPEG-2 understands this and encodes a frame in two fields, top (odd) and bottom (even).
Each field can be of type I, P, or B, with the following restrictions:
Slices
In MPEG-2, you can identify a slice, a part of a field that is guaranteed to be intracoded. Therefore, if you know that a certain region of all frames is intracoded and that no prediction is required to decode it, you can use this slice to display frames while fast-forwarding and also for reverse-playback.
Next time, we will cover the MPEG system layer.
MPEG-2 System Layer
There are three primary elements in an MPEG-2 stream:
TS packets are fixed to 188 bytes in size and each has a PID (packet ID). Once you have these packets, the decoder must be able to determine which packets contain which programs. There are two tables defined in order to do this in MPEG-2: