In general, many compressed digital sources can be transcoded to
streaming formats, but not all compressed formats transcode well. This
is due to the fact that some compressed digital formats also use
"lossey" compression methods to achieve higher data transfer rates. When
these compressed files are transcoded, areas that have had the original
data discarded are recompressed. The result is called an artifact, which
is an unnatural spot in the image that is visually distracting. MPEG1,
MPEG2, DV and compressed AVIs are all formats that can be transcoded.
The process is identical to encoding except the capture phase is not
The Encoding Triangle
As mentioned earlier, the most challenging task in encoding is balancing
between a low enough delivery rate to deliver continuously over the
target delivery rate and acceptable quality. To achieve this, you must
think in terms of a triangle. The target delivery rate is the triangle
that is fixed. Priority can then be put on any two corners at the
expense of the third. For instance, if your content is fast motion, then
image quality and frame rate are going to be important to render
acceptable action. With emphasis on these two, the frame size will have
to be lowered to keep the amount of overall data required to render the
video. If the content is a talking head, with little to no motion, then
you can increase the available bandwidth for the frame size and quality
by lowering the frame rate. In addition to this main rule, there are two
- Spacing of "key frames"The fewer key frames in the video,
the less overall bandwidth it will need to stream.
- You can also adjust the amount of bandwidth allocated to the
audio to increase available bandwidth to the video.
In conclusion, encoding is a constant balancing of settings to achieve
an acceptable result. With practice and experience, media can be produced
to dazzle all Streaming Media users on the Internet.