GstRTMPMetadata FAQ

From RidgeRun Developer Wiki

Follow Us On Twitter LinkedIn Email Share this page






NVIDIA partner logo NXP partner logo





Frequently Asked Questions


1. What is GstRTMPMetadata?

GstRTMPMetadata is not a standalone plugin but a set of patches to GStreamer’s existing flvmux and flvdemux elements. These patches extend FLV support to allow insertion and extraction of arbitrary metadata within RTMP streams, by leveraging FLV Script Data tags.

2. Which elements are modified by the patches?

The following GStreamer elements are extended:

  • flvmux: gains new properties (meta-string, meta-binary) and support for inserting metadata through GstMeta.
  • flvdemux: gains properties (attach-flvmeta, flv-meta-signal) and a new flv-meta signal to notify applications when metadata is extracted.

3. How is metadata inserted into the stream?

Metadata can be inserted in three ways:

  • As a string using the meta-string property.
  • As a binary payload using the meta-binary property (accepts a GBytes).
  • By attaching a custom GstFlvMeta to a buffer upstream, which flvmux will serialize into the FLV stream.

4. How is metadata extracted on the receiver side?

When flvdemux detects Script Data tags carrying metadata:

  • It can attach a GstFlvMeta to the first audio and video buffers if attach-flvmeta=true.
  • It can emit the flv-meta signal with the payload as GBytes if flv-meta-signal=true is enabled.

5. Which GStreamer versions are supported?

Patches are provided for:

  • GStreamer 1.16.x
  • GStreamer 1.20.x
  • GStreamer 1.24.x

6. Can metadata be updated during a live stream?

Yes. Both meta-string and meta-binary can be set multiple times while the pipeline is PLAYING. Each update results in a new Script Data tag being injected into the RTMP stream.

7. Is there an evaluation mode?

Yes. RidgeRun provides an Evaluation mode so that developers can test the product before full integration.

8. What are typical applications of GstRTMPMetadata?

  • Live telemetry synchronized with video (robotics, drones, industrial inspection).
  • Embedding timestamps, sensor fusion data, or analytics metadata into live RTMP feeds.
  • Integrations where metadata must travel alongside video in a single RTMP stream.

9. I am interested in RidgeRun services.

Please consider the following: