RTSP 1.0 vs. RTSP 2.0

Question

The Real-Time Streaming Protocol (RTSP) Version 1.0 was published as RFC 2326 in 1998.

Now nearly 20 years later Version 2.0 was published as RFC 7826 in December 2016.

I am wondering whether the changes affect the performance of live streaming using RTSP (over the Real-Time Transport Protocol (RTP)).

I know that RTSP is not used to send the real-time data, but used for session establishment and controlling mechanisms like playing, pausing or stoping the stream. So I guess the changes don't have an affect on the end-to-end latency between sender and receiver?

But in the changes it states for example

request pipelining for quick session start-up;

So my question: Is there an measurable impact on the performance regarding the introduced changes?

For example:

• session start-up time (time till the stream starts playing)
• end-to-end latency
• RTSP traffic amount
• ...

Answer 1

It depends on what your implementation supports today... if you read the news groups associated or even the first few paragraphs of the RFC you will quickly begin to understand this...

In short I believe more than having a measurable impact on performance there are changes which should hopefully create better interoperability however this is as of yet to be seen.

Most of the changes (oddly enough) are for creating and playing archived media and how to cope with such changes in the transport layer as when unsupported bandwidth meets a requested playback rate....

The most useful changes are probably the definition of the text/parameters content type and the Accept header semantics.

Pipelining is now just more widely supported and may have already been supported... IPV6 has not changed... NAT was handled better and UDP support was dropped and another type of TCP transport was supported without frame headers...

Overall though.... there's nothing else which makes Rtsp 2 any better then 1....