Using nv3dsink in a pipeline with nvcompositor

I have the following pipeline that displays the video from an RTSP camera with a digital zoom in the top-right corner:

gst-launch-1.0 rtspsrc location=rtsp://admin:123456@192.168.1.118:554/streaming/channels/101 latency=0 ! rtph265depay ! h265parse ! nvv4l2decoder ! nvvidconv ! 'video/x-raw',format=RGBA ! queue ! tee name=t \
t. ! queue ! comp.sink_0 \
t. ! queue ! videocrop top=315 left=520 right=520 bottom=315 ! videoscale ! comp.sink_1 \
nvcompositor name=comp sink_0::xpos=0 sink_0::ypos=0 sink_0::width=1280 sink_0::height=720 sink_1::xpos=960 sink_1::ypos=0 sink_1::width=320 sink_1::height=180 ! nvoverlaysink

This pipeline works fine with nvoverlaysink, but when I try to display the video in a GTK window using GstVideoOverlay, nvoverlaysink does not work.

I tested the following pipeline with nv3dsink, and it displays the video correctly in the GTK window:

gst-launch-1.0 rtspsrc location=rtsp://admin:123456@192.168.1.118:554/streaming/channels/101 latency=0 ! rtph265depay ! h265parse ! nvv4l2decoder ! nvvidconv ! 'video/x-raw(memory:NVMM)' ! nv3dsink -e

I would like to know how to use nv3dsink in the first pipeline, while still keeping the video composition with nvcompositor. Any suggestions?

Hi,
Please refer to the pipeline and give it a try:
Nvcompositor can not connect nv3dsink when nvcompositor sinkpad is system memory - #5 by DaneLLL

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.