Pixel format conversion using VIC


We are developing some hardware based on the Tegra K1 and we are interested in using the Video Image Compositor (VIC) described in the TRM (section 13). I would like to know if there are some examples how to use the features described in the documentation. We are specially interested in using color conversion (YUV2 to RGBx) and scaling.

Here is multimedia user guide covering gstreamer using some of the VIC functions,

I also want to know if there are more details about VIC. In other words, how can I program it? Especially, the Blend/Composite features which is shown in TRM(section 13)?

bump. Any updates regarding this topic from nvidia? Is it possible to use the VIC directly for processing data, rather than rely on the gstreamer plugins? I find that the HW accelerated gstreamer plugins are slower than the other software plugins available.

The HW accelerated GST plugins are significantly faster than the SW plugins for the common operations. E.g. you can’t decode 2160p30 without the HW plugins.

It is possible to ruin the performance benefits by doing e.g. extra data conversions. And in some cases some data conversions can’t be avoided.

This sub-forum is for the older SOCs. You may want to use the Jetson TK1 subforum to ask more details related to your GST performance issues. GStreamer is the official multimedia API for Jetsons so trying first to solve the issues with that might be the fastest approach.

Thank you for the info kulve. Didnt see that this was actually a sub-forum for the older SOCs. Just read the first post and assumed this to be Jetson TK1 thread.