I have some questions about the video Port:
As described in "spru629" the video Port is capable among others of
acquiring frames, perform scaling and chrominance resampling.
In the configurations of the video port, the threshold to perform a
DMA transfer for the chrominance components is automatically defined
as half of the threshold defined for the luminance component.
My questions are concerned about the performance of the video Port, as
well as acquiring frames in resolutions others than 4:2:2.
1. The video Port is capable of performing scaling as well as
chrominance resampling. This operation is "computationally
demanding"??? The DSP have any intervention on this??
2. Having a video decoder capable of performing scaling and
chrominance resampling, i think this is preferable instead of this
being done by the Video Port. Nonetheless one question arise. Imagine
that we want to capture images in 4:2:0 resolution. In this case each
of the chrominance components have 1/4 of the size of the luminance.
Since the chrominance threshold for DMA transfers is automatically
defined by the threshold of the luminance component, how can we
capture images in 4:2:0 resolution??
Thanks in advance,