About gops (groups of pictures) – Apple Compressor (4.0) User Manual
Page 99
Chapter 5
Custom settings and output formats
99
About GOPs (groups of pictures)
A major feature of MPEG-2 encoding is its ability to remove redundancy, not only within a frame
but also among a group of frames. MPEG-2 uses three frame types (I, P, and B) to represent the
video. A group of pictures (GOP) setting defines the pattern of the three frame types used. These
three frame types are defined in the following ways:
•
Intra (I): Also known as the key frame. Every GOP contains one I-frame. The I-frame is the only
MPEG-2 frame that can be fully decompressed without any reference to frames that precede
or follow it. It’s also the most data-heavy, requiring the most disk space. If you want to place an
I-frame at a scene change or some other specific frame location, you need to manually set it
using the Preview window. This is known as a forced I-frame. For more information, see
on page 46.
•
Predicted (P): Encoded from a “predicted” picture based on the closest preceding I- or P-frame.
P-frames typically require much less disk space than do I-frames because they reference a
preceding I- or P-frame in the GOP.
Note: Both I-frames and P-frames are also known as reference frames, because a B-frame may
refer to either one or both frame types.
•
Bi-directional (B): Encoded from an interpolation of succeeding and preceding reference
frames, either I-frame or P-frame. B-frames are the most storage-efficient MPEG-2 frame type,
requiring the least amount of disk space.
The use of B- and P-frames is what allows MPEG-2 to remove temporal redundancy, contributing
to its ability to compress video efficiently.
You need to consider the following factors when choosing a GOP setting.
GOP structure
This setting specifies whether there will be two, one, or no B-frames between the reference
frames within a GOP. GOP structure, along with GOP size, determines the number of I-, P-, and
B-frames that are used during transcoding.
The GOP structure you choose depends on how far apart P-frames should be spaced. Because a
P-frame is predicted from the previous reference frame (either an I-frame or a P-frame), if there
are one or two B-frames in between, the prediction must cover the distance objects can move
over the duration of two to three frames.
In principle, the less average motion there is from one frame to the next, the farther apart
P-frames can be spaced, and the greater the compression can be. For most video material,
the IBBP structure is a good choice. Material with unusually fast motion throughout the entire
sequence may benefit from an IBP or IP structure, but in such cases a relatively high bit rate (6 to
8 Mbps for SD video) may be required for good quality.
GOP size
This setting specifies the number of frames within a GOP. Because exactly one I-frame exists
per GOP, longer GOP sizes generally provide greater compression, because B- and P-frames are
smaller than I-frames.
For most media, spacing I-frames about 1/2 second apart gives good results. This equates to a
GOP size of 15 frames for NTSC and 12 frames for PAL. The DVD-Video specification prohibits GOP
lengths from being much longer than this. Generally, only material with scene changes occurring
less than 1/2 second apart frequently throughout the video benefits from shorter GOP sizes.