Latency
Latency is the delay between something happening in front of a camera and a viewer seeing it on their screen. If you hear your neighbor cheer a goal before you see it on your own TV, the gap you are noticing is latency. In live streaming it is measured in seconds, from capture to playback.
Why it matters
Latency decides whether live video feels live. High delay spoils results through social media, puts chat and polls out of step and makes betting or bidding unfair. For on-demand video, startup time matters more, but for live products latency is part of the experience being sold.
How it works
- The camera captures the scene, and the encoder compresses it, adding a small delay.
- The stream travels to the server, which transcodes and packages it into segments.
- A CDN delivers those segments to the viewer's device.
- The player keeps a buffer before showing video, which often adds the largest share of delay.
What is a good latency for live streaming?
A good latency for live streaming is about 2 to 5 seconds for most sports, events and live shopping. Standard HLS often runs 10 to 30 seconds behind, which is fine for one-way broadcasts. Two-way conversation, betting and auctions need under one second.
How can I reduce latency in a live stream?
To reduce latency in a live stream, switch to a low-latency protocol such as LL-HLS, use shorter segments and keyframe intervals and lower the player's buffer. For under one second, use WebRTC. Each step trades some stability for speed, so test on real networks. FastPix delivers every live stream as LL-HLS, so viewers stay closer to the live moment without your team rebuilding its delivery stack.
Latency vs buffering: what is the difference?
Latency is how far behind real time a stream is, while buffering is when playback stops to wait for more data. A stream can have high latency and never buffer. Cutting latency too far can raise buffering, because the player has less data in reserve.
Where it's used
Live sports streamers, live shopping apps and fitness classes keep latency low so fans, buyers and riders react in the moment. On-demand libraries focus on startup time instead.
Where it usually goes wrong, and how FastPix handles it
Teams often guess their stream delay from a single test instead of measuring what real viewers see across devices, networks and regions.
Live stream latency for each viewer is recorded in FastPix Video Data, with breakdowns by country, device and CDN. The region or device falling behind shows up before viewers start complaining.
By the numbers
- YouTube's Low latency setting gives most viewers under 10 seconds of delay, and its Ultra-low latency setting gives most viewers under 5 seconds. (Understand live streaming latency (YouTube Help))
- The HLS specification (RFC 8216) gives 10 seconds as a typical segment target duration. (RFC 8216: HTTP Live Streaming (IETF))
