MTD2 class 13

esse quam videri
Revision as of 22:19, 29 November 2007 by Jeff (talk | contribs) (Links)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

In class

  2. Video Compression
  3. Compression in Premiere
  4. Continue work on Animation

Image Compression

JPEG determines for an image is turned into a byte stream. Converts from RBG to YCbCr JPEG

lossy image compression

PNG

lossless image compression

critical and ancillary. ancillary chunks are used to extend png capabilities while maintaining compatibility with older versions. Ancillary chunks my safely ignored.

PNG is also paren free and based on open source library libpng Discrete cosine transform

MPEG

MPEG Overview

Temporal Redundancy (inter-frame)- successive frames contain the same information

Spatial Redundancy (intra-frame)- redundancy within a single frame

MPEG2 Three way on encoding a frame

(I)intra-frame
(P)forward predictive frames
(B)bidirectional predictive frames

I frames contain a full compressed images of the frame (Spatial). P frames contain predictive information to help remove Temporal redundancy. Frames are compared through time and through a process called "motion compensation" a motion vector map is created and only the differences between frames is coded. P frames only work in the forward direction. B frames are encoded against the next and previous frames and work in both directions. P and B frames are usually 90%-98% smaller than I frames.

These frames are arranged into a GOP (group of pictures). There are many possible structures for each GOP a common 15 frame GOP could be structures like

I_BB_P_BB_P_BB_P_BB_P_BB

DiagramMPEG.png

MPEG Artifacts
codec support

Homework

Continue work on flash animation

Premiere Titles and Fades

Links

Beginners guide to MPEG-2 Standard
MPEG-2 Compression MPEG Artifacts webflv