Multipass Encoding

Multipass encoding converts video files.

Multipass encoding is a system for converting one video format into another. It is also known as 2-pass encoding because the program encoding the video processes the source file twice.

Function

The first pass in a multipass conversion analyzes the frames in a video to assess the complexity of its contents. For example, a frame showing all one color needs little clarity so the conversion program reduces the bit rate expended on these frames to make greater bit rate available to encode more complicated frames. This gives greater clarity to the video overall as resources are focused on certain frames. During the first phase, the program writes notes to a log file. These inform the encoding phase which is the second pass of the conversion.

Features

Multipass encoding is closely associated with Variable Bitrate encoding (VBR). VBR is the strategy of storing video frames in a file in different amounts of bits. The alternative to VBR is Constant Bitrate encoding (CBR) which uses the same number of bits for each frame. CBR does not require multipass encoding.

Implementation

DivX versions 4 and 5 offer multipass encoding options. The XviD, Gordian Knot, AutoGK and SUPER encoding systems also have a multipass option. There is no specific open standard defining the protocol.

References

• AfterDawn: Multi-pass encoding
• AfterDawn: VBR
• Digital Digest: DVD to DivX (Xvid) Conversion Guide