Chroma sub-sampling converts the video to 4:2:2 format by horizontally sub-sampling the Cb and Cr components by a factor of 2. This is a simple DSP sample rate conversion operation, in which the new sample rate is exactly half the old sample rate. Up-sampling is the process of converting 4:2:2 format back to 4:4:4 by over-sampling the chroma components by a factor of 2.
The video rendering pipeline contains an up-scaling color depth scaling block at the input and down-scaling color depth with dithering at the output. Depending on the configuration, the pixel scaling block can scale a lower bit-per-color (BPC) value to a higher BPC value. The pixel descaling block can convert a higher BPC to a lower BPC at the output of the video blender block. The pixel scaling block converts low-resolution pixels to high-resolution pixels after multiplying the input padded pixels by a scale factor. Pixel descaling is done at the video blender output. Dithering reduces the contouring artifacts that occur at low pixel resolutions. The dithering operation consists of the following.
•Add a dithering value
•Saturate
•Truncate to the desired size
The video blender (This Figure) takes two native video stream inputs and outputs blended pixels. The blending operation converts two input streams into RGB. The final result is converted to proper format per user-supplied programming. The final output video is also forwarded to the PL layer.
