Error latency values for gstreamer pipelines using latency-clock plugins for two PCs in the LAN

Viewed 151

I tried to stream the camera feed from my Ubuntu 20.04 installed Khadas vim3 to my Ubuntu 16.04 installed PC. Those two nodes are connected in the same wired LAN. My stream was successful. After that, I wanted to measure the latency value quantitatively. For that, I used latency-clock as gstreamer plugins and tried it out. Commands as follows,

Streamer (Khadas vim3)

gst-launch-1.0 -v --gst-plugin-path={YOUR PATH}/latency-clock v4l2src device=/dev/video0 ! video/x-raw,format=RGB,framerate=30/1,width=1920,height=1080 ! timeoverlay ! timestampoverlay ! videoscale ! videoconvert n-threads=1 ! x264enc tune=zerolatency bitrate=2048 speed-preset=superfast ! rtph264pay ! udpsink host={HOST_IP} port={PORT}

Receiver (PC)

gst-launch-1.0 --gst-debug=timeoverlayparse:6 --gst-plugin-path={YOUR PATH}/latency-clock -v udpsrc port={PORT} caps = "application/x-rtp, media=(string)video, clock-rate=(int)90000, encoding-name=(string)H264, payload=(int)96" ! rtph264depay ! avdec_h264 ! videoconvert ! timeoverlayparse ! fakesink sync=true

All are worked properly. But on the receiver side, in debug output, the latency reading is an abnormally large value. I tried with the NTP server and synchronized the time of the PC with the Khadas SBC and there is no change with the error.

Debug Output

Output

As well as if the streamer and the receiver run on the same PC, the latency values are completely ok. But in the different machines in the LAN, the latency values are abnormal. Could anyone please suggest a solution to this issue? Thankyou!

0 Answers
Related