Appendix: Advanced Settings for H.264

Setting

Explanation

 

 

 

Because teas (10) is not achievable in real time encoding there

 

is no reason to use it.

 

dia – (diamond) is the simplest search. It starts at the best

 

predictor, checks the motion vectors at one pixel upwards,

 

left, down, and to the right, chooses the best, and repeats the

 

process until it no longer finds any better motion vectors. This

 

setting is the fastest.

 

hex – (hexagon) is the default setting. It uses a similar strategy

 

to dia, except it uses a range-2 search of 6 surrounding points.

 

It is considerably more efficient than dia and hardly any

 

slower. This setting is a good choice for general use.

 

umh – (uneven multi-hex) is considerably slower than hex. It

 

searches a complex multi-hexagon pattern to avoid missing

 

harder-to-find motion vectors. umh is about 40 % slower than

 

hex.

 

esa – (exhaustive) is a highly optimized intelligent search of

 

the entire motion search space within range of the best

 

predictor. It is the mathematically equivalent to the brute

 

force method of searching every single motion vector in that

 

area, though faster. It is still considerably slower than umh.

 

teas – (transformed exhaustive) attempts to approximate the

 

effect of running a Hadamard transform comparison at each

 

motion vector. It’s a little bit better than esa but a little bit

 

slower, too slow for practical use.

MotionEstimationRange

This value controls the maximum range of the motion search.

 

For hex and dia, the encoder only uses between 4 and 16. umh

 

and esa allow you to increase past 16 to allow a wider range of

 

motion search, which is useful for high-definition footage and

 

for high-motion footage. Increasing MotionEstimationRange

 

significantly slows down encoding.

MotionEstimationSubPelQuality

This value controls the subpixel motion estimation quality.

 

Higher numbers are better. Levels 1 through 5 simply control

 

the subpixel refinement strength. Level 6 enables RDO for

 

mode decision, and level 8 enables RDO for motion vectors

 

and intra prediction modes. RDO levels are significantly slower

 

than the previous levels.

 

0 — fullpel only (not recommended)

 

1 — QPel SAD 1 iteration

 

2 — QPel SATD 2 iterations

 

3 — HPel on MB then Qpel

 

4 — Always Qpel

 

5 — Multi QPel + bime

84

ViewCast

Page 94
Image 94
ViewCast 2120 manual Is no reason to use it