XviD Technology _ Optimization

Multiple encoding modes:

In addition to the most original single revaluation stream compression (1-pass CBR), XviD includes two dual (2-pass) dynamic code stream compression modes including single weight mass mode dynamic code flow compression, single weighting (quantization) mode dynamic code flow compression, and both external control and internal control 。 XviD is clearly the most MPEG4 coding core in the current PC, with the most optional mode of video coding.

In particular, the dual (2-pass) dynamic code stream compression mode, through the detection of the screen motion amplitude in the most optimized way to distribute the picture quantization amplitude, so full of active elements of the video film can occupy space and the quality of the picture to achieve the best balance. and the single weight (quantization) mode dynamic code flow compression, can be generated at a high speed control screen details of the dynamic code stream video files, in less than the compression time cost, you can get a more stable picture detail quality, this is simple DivX 3.11 does not have excellent features.

Quantification (quantization) mode and range control

Quantification (quantization), simply is in the encoding of the time or space adjacent to the picture unit assimilation, fuzzy details of the degree, is the screen quality of the most basic control factors. XviD not only provides the standard MPEG quantization method, but also specially provides the h263 quantization method which is more suitable for the low code flow compression. And XviD can also be in the double (2-pass) operation, according to the picture of the comprehensive analysis of information, dynamic decision of a scene of the picture quantization, it can be said to be the network media file dissemination, space and quality balance and consideration, design intimate got home.

In addition to the quantification options, XviD also provides a powerful range control over the quantization amplitude in the compression process. Users can select a range of quantization ranges to allow for compression. For example, if you set a quantitative limit, you can avoid the possibility of a sharp decline in quality.

Motion detection (Motion Search) and curve balance allocation (Curve)

Motion detection by frame on the screen, after analyzing the motion detection result of the whole fragment, the quantization amplitude of each frame is allocated by the curve balance to achieve the following: The motion picture needing high code flow can allocate more space, higher code flow and lower quantization amplitude to keep the detail of the picture. For static images that do not contain too much motion information, the allocation budget is reduced. The use of good steel in the cutting edge of the practice, is XviD as the second generation of MPEG4 coding core content.

The XviD provides a multipole motion detection precision, including a half interpolation (pixel interpolation) technique, which is labeled as a motion vector by a cell of 16x16 pixel; And a 4-point motion vector (inter4v motion vectors), the 8x8 pixel block is a more detailed recording motion vector for double analysis.

Dynamic Keyframe distance (i-frame interval)

Dynamic keyframe spacing is one of the techniques that another XviD has to achieve maximum balance between space and picture. We know that in the video compression is not every frame is recorded in all the picture information, in fact, only the key frame records the complete picture information, and subsequent P frame (p-frame) is only a record of the difference between the previous frame. If the picture between the key frame changes very big, then will waste the precious space on the p-frame; The addition of a large change in the frame recorded in the key frame, then because the subsequent frames no longer have a greater change, you can save the space required by P frames. Therefore, according to the scene lens switching and motion amplitude to transform the location of key frames, video compression under the image quality improvement, there is a multiplier effect.

Brightness correction of mental vision

In addition to the basic MPEG4 coding, XviD also offers a number of additional features. One of the typical is mental visual brightness correction. This function may increase the compression efficiency by removing the luminance signal and brightness difference that cannot be distinguished by the naked eye. Unfortunately, although the design is very creative, but the actual application of the current will lead to the visual quality of the image drop, but also look forward to the future version can be improved.

Credits options

Another intimate design is a movie-specific "cast-list option." This feature allows for a very low stream of code to be set in a user-specified paragraph where the detail is not necessary (for example, a movie's credits). Even compressed to the 10% of positive code flow, you can save a lot of space, the design is also very intimate.

External custom Control

In addition to XIVD own internal curve allocation control mode, XviD also provides an external open interface. Allows users to skip the core of XviD's own coding analysis, using external analysis tools provided by third parties, such as Gordian Knot, to generate allocated control files, and then return XviD to do the final double operation compression. This approach extends the usability and user control of XviD, as Gordian Knot software can even be controlled by the user to control the amount of quantization and stream allocations per frame, providing more possibilities for advanced applications.

Motion vector Acceleration (hinted Me) encoding

XviD also has a design designed to improve the coding efficiency, which is to record the motion vector information of the picture into a Log file. Secondly, by reading the information of the file directly, the repetitive operation of the motion vector information is saved and the coding speed is greatly improved.

Picture optimization decoding

Not only on the code XIVD has a powerful function, in decoding the picture optimization, XviD also has a lot of new achievements. For example, "Horizontal Deblock (Y)" can soften the horizontal direction of the brightness mosaic; "Horizontal Deblock (C)" Can soften the color mosaic of the horizontal phase; "Deringing (Y)" can soften the ring brightness spot; "Deinging (C)" Can soften the ring chroma color spots; "Mosquito" can reduce the mosquito-like spots in the screen. Also, these XviD are optional, so you can play them on a lower-configured PC as long as you turn off several less-than-necessary options.

B-Shadow being developed

Another notable technological development in the XviD is the B-shadow. We know that the I shadow is the key shadow, the P shadow between the key, only the difference between the previous uncompressed picture. B-Shadow and the principle of the P-shadow, but in addition to the three extracted before the screen, will also refer to the next shadow of the picture information. So the B-shadow extract the picture is better than the P-shadow will come.

The previous MPEG4 coding cores used only the I and P shadow storage images. If you can use B-shadow technology on this basis, nature can be a higher level in quality and compression ratio. At present, XviD has provided developers with the test of the nature of the beta XviD coding core, I believe that in the near future, XviD will be able to play the advantages of MPEG4 coding to the extreme, become the online video recording format of the leader.