Bit Allocation

Tarkin-Dev Contributors,
When it comes to lossy compression, an efficient allocation of bits is
extremely important. Just as when packing a suitcase, you must decide what
to put in and how to arrange it (bit allocation), before eventually cramming
the lid down (compression) :).

The eye is actually a very poor optical device, in terms of resolution. The
human retina is divided into three parts: a low resolution area used for
peripheral vision; the Macula, a medium resolution area for seeing shapes;
and the fovea, a very very small, very very high resolution area for tasks
such as reading. When confronted with a frame of video to examine, the eye
darts around, focusing areas of interest on the fovea. Since each video shot
doesn't last very long, the fovea can only appreciate a few key details
before it is presented with a new image to examine.

Facial features like eyes and mouths are followed closely by the fovea,
along with fast moving objects and objects deemed important by the brain.
Therefore, these areas should recieve more bits than others.

On with my idea :). After motion compensation is done, the codec analyses
the MVs and finds blocks/areas which are moving in a different direction to
the ones around them. These areas which are moving "against the grain" are
allocated more bits since they are more interesting? Any ideas...

Dust

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to '[email protected]'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.