CMAF technology. Low Latency Streaming

To configure live streams delivery using Low Latency MPEG-DASH, contact technical support

What is the stream latency?
Stream latency reasons
Low Latency MPEG-DASH
How Low Latency MPEG-DASH works

What is stream latency?

Stream latency is a time interval between the moment a frame is captured and the moment that frame is displayed on the viewers' screens.

The average stream latency is 30-40 seconds.


Stream latency reasons

A video stream is delivered to end-users through several stages:

  • Encoding/packaging: content is being processed to create stream segments.
  • Transfer: upload to a CDN and deliver to the user's playback device.
  • Players: segments loading and buffering them.

Each stream conversion stage increases latency, which can be critical for certain types of content: sports events, news releases, events related to esports, and interactive.

Low Latency MPEG-DASH

MPEG-DASH (from MPEG and Dynamic Adaptive Streaming over HTTP) is an adaptive data streaming technology for delivering streaming media content over the Internet via the HTTP protocol.

CMAF (Common Media Application Format) is a tool that allows reducing the size of the chunks. 

Low Latency MPEG-DASH is a video delivery technology that allows low-latency due to the use of the MPEG-DASH protocol and CMAF technology.

This method reduces latency to 4-5 seconds.

How Low Latency MPEG-DASH works

The LL-DASH mechanism uses an extension of the CMAF - CMAF-CTE (Chunked Transfer Encoding) technology.

A standard method of video delivery is when the encoder sends the segment to the CDN after its complete creation and as soon as the CDN receives the full segment, it passes it to the player.

With this approach, the video latency depends on segment length.

For example, if a segment is 7 seconds long, when requesting and processing the first segment, the player displays a frame that is already 7 seconds late compared to the actual time.


Using CMAF-CTE, LL-DASH segments are divided into chunks — smaller non-overlapping and independent of one another segments of 2-4 seconds.

The independence of chunks allows the encoder not to wait for the end of the full loading of the segment, but to send it to the CDN and to the player in ready-made small fragments.

This approach helps to eliminate the segment duration factor that in standard video delivery method affects the video delay.

This means that for a segment with a length of 10 seconds and a segment with a length of 2 seconds, the latency will be the same.  

The total latency between the content delivery server and the end viewers is no more than 4 seconds.


To use the Low Latency MPEG-DASH technology, it is necessary that CMAF is supported at all stages of video transmission, so we recommend using a player that supports the dash.js library.

Use your own or the Gcore player.

Important! iOS's devices don't support dash.js, so for our player we use HLS fallback.

If you are going to use a different player, make sure that the necessary settings are made.

To check the Low latency, go to the demonstration page.


Was this article helpful?
Recently viewed articles