Today i restarted with GStreamer after around 2 months. I started using GStreamer to generate RTP traffic, capture it via Wireshark and save it as ASCII file. My first two blogs talk about these. Since we are dealing with uncompressed video, using GStreamer to generate RTP traffic didn't work.
Now the goal is to use GStreamer to reconstruct video frame from RTP/UDP traffic coming at the Ethernet interface. From some discussion at IRC channel #gstreamer, it seems it is more than creating a simple Gstreamer pipeline. I have been advised to read up GStreamer Application Development manual, which i am doing right now. From reading a few chapters, it seems i have to write a C program to get the job done. But here and there i see Gstreamer one-liners as well. I am going to check it tomorrow on #gstreamer. If you can suggest something, please do. I am learning the ropes when it comes to Gstreamer.
I missed the IRC chats (link to logs?), you might want to guesstimate how much effort it will take to get a gstreamer environment up and running. If it looks like a week or so we might want to consider other verification routes.
ReplyDeleteIf you are unable to use an existing framework you can write a basic Python application (similar to the one used in the early verification) to strip out the data and create an image (e.g. using PIL).
I know absolutely nothing about gstreamer I will not be of much use. But I want to make sure the focus is on verifying and documenting the RTP video stream packetizer, which ever method is the least resistance to fully verify I support.