Skip to content
  • There are no suggestions because the search field is empty.

Introduction to SRT

SRT (Secure Reliable Transport) is a streaming protocol designed for resilient, dependable streaming from Point A to Point B.

 

This is typically a peer-to-peer streaming method and is not intended for multiple destinations and viewers to access. Because of this, once an SRT connection has been established, it will only work for one encoder-decoder pair.

 

Secure

SRT streams can be secured using AES encryption. This will protect your stream from being viewed by unwanted parties. This security can also keep your stream from being interfered with by outside parties.

Reliable

Typically, if a segment arrives incomplete or partially complete via RTMP or HLS, then the player would buffer, unable to operate until the next segment arrives. SRT streams have an opportunity to repair these incomplete segments, or deliver missing segments depending on how much latency is used in the stream. Increasing a stream’s latency gives the encoder and decoder time to work together on addressing problematic segments. 

 

Typical Use Cases

  • Streaming a single-camera feed or program mix from event venue to a master control facility
  • Streaming from a master control facility to a broadcast TV network
  • Sending a program mix or camera feed from one venue to another venue

 

Caller and Listener

In normal streaming workflows, we are used to common terms such as encoders, decoders, codecs, sources and destinations. When we think of encoders and sources, we assume that this is where our stream is starting. Conversely, when we think of decoders and destinations, we assume that this is where our stream is ending.

SRT introduces two new terms to streaming workflows, caller and listener. These terms are not mutually exclusive, nor do they fit into our normal assumptions with start and end points.

Caller and listener do not relate at all to the actual streaming workflow. Instead, they are responsible for establishing and maintaining the peer-to-peer connection. This means that the decoder can be the caller, and the encoder the listener. Or vice-versa. 

In a typical SRT workflow, there is one SRT caller, and one SRT listener. One of these must be the caller, and the other the listener.

  • A listener sits at specific IP address (or URL) and is monitoring a port. 
  • A caller can be anywhere in the world, and it “calls” the listener at its IP address (or URL) and its port.

The caller will seek out the listener. Once found, the listener can ask the caller for a passphrase, protected by AES 128- to 256-bit encryption. The listener will validate the passphrase and accept the connection. 

With the connection established, the encoder and decoder will perform their expected functions and the peer-to-peer streaming will begin.