Live stream latency explained
Latency is the delay between something happening in front of your camera and a viewer seeing it. A little is unavoidable; how much you want depends on whether you need real-time interaction or rock-solid stability. Here’s where the delay comes from and how the tiers compare.
Where the delay comes from
Latency adds up across the chain: your encoder’s buffer, the upload and network hops, the ingest server, the platform’s transcoding into multiple qualities, delivery over a CDN, and finally the player’s own buffer on the viewer’s device. The player buffer and platform transcoding are usually the biggest contributors.
The latency tiers
Standard latency (≈15–30s) is the most stable and compatible, and is fine for most one-way broadcasts. Low-latency (≈3–10s) keeps live chat feeling responsive and suits most creators. Ultra-low / WebRTC (under a second) enables true real-time interaction like auctions or two-way Q&A.
The trade-off: latency vs stability
Lower latency means smaller buffers, which leaves less room to absorb a network hiccup — so very low latency can mean more stuttering on a shaky connection. Pick the lowest tier your use case actually needs, not the lowest possible, and keep upload headroom so the stream stays smooth.
Frequently asked questions
What is a normal live stream delay?
Most platforms run around 15–30 seconds by default, with optional low-latency modes around 3–10 seconds.
Does restreaming add a lot of latency?
Very little — a copy-first relay passes your encode through without re-encoding, so it adds minimal delay on top of the platforms’ own pipelines.
Why is lower latency sometimes worse?
Lower latency uses smaller buffers, which leaves less room to ride out network dips — so it can stutter more on an unstable connection.