Skip to main content

​
What is GStreamer?

GStreamer is an open‑source multimedia framework used to build streaming pipelines that capture, process and transmit audio and video. Instead of being a single application, GStreamer provides a modular set of elements that can be assembled to perform tasks like media playback, recording, transcoding or live broadcasting. This flexibility makes it popular for both desktop applications and server‑side streaming platforms, including broadcasting software for live events. For live streaming, GStreamer can act as the “encoder” component of your broadcast workflow. It takes raw video/audio from sources (e.g. cameras or test patterns), encodes them with a codec (such as H.264 for video and AAC for audio), packs the streams into a container and sends them to an ingest server over protocols like RTMP or SRT. An example pipeline below is a concise example of how to contribute a stable test stream to the Gcore RTMP/SRT ingest point using GStreamer:
  • Synthetic test sources: videotestsrc pattern=smpte and audiotestsrc wave=sine produce a continuous 1080p, 30 fps color bar video and a sine‑tone audio track. This allows you to test the ingest without needing any external media files.
  • Efficient, low‑latency encoding: The caps filter enforces YUV 4:2:0 (I420) and 30 fps, while x264enc tune=zerolatency combined with key‑int‑max=30 and bframes=0 produces a GOP every second and avoids B‑frames. This keeps latency low and ensures the stream is easy for downstream transcoding.

​
Pushing via RTMP

To push a test stream via RTMP, you can generate a synthetic video and audio source, encode them, wrap them into an FLV container and send them to your ingest endpoint. The following command demonstrates how to do this with GStreamer: 
gst-launch-1.0 -v \
    videotestsrc is-live=true pattern=smpte ! \
        video/x-raw,format=I420,width=1920,height=1080,framerate=30/1 ! \
        x264enc tune=zerolatency bitrate=3000 speed-preset=veryfast key-int-max=30 bframes=0 ! \
        h264parse config-interval=1 ! queue ! mux. \
    audiotestsrc is-live=true wave=sine ! \
        audio/x-raw,channels=2,rate=48000 ! \
        avenc_aac bitrate=128000 ! queue ! mux. \
    flvmux name=mux streamable=true ! \
    rtmpsink location="rtmp://vp-push-ed2.gvideo.co/in/123?719"

​
Pushing via SRT

For stability contribution, you can push the same test stream via SRT instead of RTMP. The main differences are the container (MPEG‑TS rather than FLV) and the sink element. SRT also allows you to specify latency and mode parameters. Here is a sample SRT pipeline using the provided ingest URL: 
gst-launch-1.0 -v \
    videotestsrc is-live=true pattern=smpte ! \
        video/x-raw,format=I420,width=1920,height=1080,framerate=30/1 ! \
        x264enc tune=zerolatency bitrate=3000 speed-preset=veryfast key-int-max=60 bframes=0 ! \
        h264parse config-interval=1 ! queue ! mux. \
    audiotestsrc is-live=true wave=sine ! \
        audio/x-raw,channels=2,rate=48000 ! \
        avenc_aac bitrate=128000 ! queue ! mux. \
    mpegtsmux name=mux ! \
    srtsink uri="srt://vp-push-ed2-srt.gvideo.co:5001" \
            mode=caller latency=1500 \
            streamid="123#719"
This pipeline mirrors the RTMP example but uses an MPEG‑TS container and the SRT protocol:
  • “mpegtsmux” multiplexes the H.264 video and AAC audio into a MPEG‑TS stream, which is a more common container for SRT.
  • “srtsink” sends the MPEG‑TS stream to the SRT ingest URL. The “uri” parameter is taken from UI/API from igester server field. The mode=caller option says that GStreamer initiates the connection, the latency in milliseconds (1.5 sec in this example).
  • Pay attention to the stream identification: The “streamid” parameter uniquely identifies your stream to the server. Using the uri plus separate mode and latency properties ensures the # in your stream ID is passed correctly.