Video distrabution has historically relied on RTMP. While widely supported, RTMP is a legacy protocol that struggles with modern encryption standards and lacks mechanisms to handle network instability.
Streaming protocols like HLS are robust for delivery to viewers but the inherent latency makes them unsuitable for interactive contribution. If you process live media you cannot afford a 30 second delay.
Fifth Layer now supports native SRT (Secure Reliable Transport) ingestion. You can provision dedicated ingestion endpoints and process low-latency streams using our visual workflow builder.
The Protocol
SRT is an open-source video transport protocol. It’s robust to packet loss and jitter to ensure a pristine video stream that can traverse firewalls while maintaining latency in the range of hundreds of milliseconds.
Designing the Workflow
In Fifth Layer, SRT is treated like any other operator.
You start by defining the logic in the workflow editor. Drag an SRT Input node onto the canvas. Unlike a static file input, this node waits for a signal from an external source.
Once the node is on the canvas you can connect it to the rest of your pipeline. The video is decomposed into the audio and video components, allowing you to route to any other operator in the system.
Provisioning the Endpoint
To feed data into the SRT Input node you require a Trigger.
Navigate to the Triggers section of the dashboard and select SRT Stream Listener. This instructs the Fifth Layer infrastructure to provision a unique endpoint for your workflow.
Once created the system provides a connection string. This URL contains the information required to route the video to your specific workflow.
This URL is compatible with all major hardware encoders and software such as OBS Studio or AWS Media Live.
Invocation
We do not maintain persistent servers waiting for your stream. When your encoder connects to the endpoint it triggers an Invocation. Fifth Layer instantly spins up a fresh instance of your workflow specifically for that session.
This ensures you are only billed for active ingestion. Whether you are handling a single program feed or fifty isolated camera angles we scale elastically to support your needs.
Processing Opportunities
Once the stream is inside the graph you can apply any logic required.
AI Integration - The low latency of SRT makes it ideal for real-time AI processing. You can extract audio for AWS Transcribe or send the entire asset to OpenAI and Google Gemini for immediate analysis.
Social Distribution - Use the Rescaler and Canvas nodes to reformat standard 16:9 broadcast feeds into vertical formats for social distrabution. These can be pushed directly to platforms like TikTok or Instagram.
We have removed the infrastructure complexity from SRT. We provision the listening ports and handle the handshake so you can focus on the logic of your production.