New: Enveu Flow is now generally available — automate media operations alongside Experience Cloud. Learn more
Live Delivery

Latency

Reviewed by Abhinav · OTT & StreamingLast updated: 2026-06-30

Latency in streaming is the delay between a live moment happening and a viewer seeing it. Standard HLS/DASH latency runs 15–45 seconds; low-latency and ultra-low-latency techniques cut it to a few seconds or less for interactive live.

Enveu take
Latency only matters when the audience knows what they're missing — for sports and interactive live, seconds behind the neighbour's TV is a real problem; for VOD it's irrelevant.
LiveDeliveryPerformance

Where it fits in the OTT stack

EncoderPackagerCDNPlayer Buffer

How it works

  1. Each stage — encode, package, deliver, buffer — adds delay.
  2. Segment duration and player buffer are the biggest contributors.
  3. Low-latency modes use shorter segments and chunked transfer.
  4. The player trades buffer size against how far behind live it stays.

Key components

  • Segment/chunk duration
  • Player buffer target
  • Protocol mode (LL-HLS, LL-DASH, WebRTC)
  • CDN and origin delivery path

Performance impact

  • Standard HLS/DASH: ~15–45s behind live
  • Low-latency HLS/DASH: ~2–8s
  • WebRTC/ultra-low: sub-second, at scale cost

Common issues

  • Segments too long for the target latency
  • Buffer set high, increasing delay
  • Trade-off: lower latency can raise rebuffering risk

When it matters

  • Live sports and betting
  • Auctions, interactive, and watch-along live
Signals to monitor
  • Glass-to-glass latency
  • Rebuffer ratio at low latency
  • Concurrency at peak

Real-world example

Getting sports closer to live
A sports service ran 30+ seconds behind broadcast.
Challenge
  • Viewers saw goals late via social/second screen
  • Standard segments were too long
Action taken
  • Switched to low-latency HLS with shorter chunks
  • Tuned the player buffer for the new target
Outcome
Glass-to-glass latency dropped to a few seconds, closing the gap with broadcast.

Frequently asked questions

What is latency in streaming?
Latency is the delay between a live event happening and a viewer seeing it. Standard HLS/DASH runs 15–45 seconds; low-latency methods cut it to a few seconds or less.
What is low-latency streaming?
Low-latency streaming uses shorter segments/chunks and reduced buffering (e.g. LL-HLS, LL-DASH, WebRTC) to bring the delay down to a few seconds or sub-second for interactive live.
Does latency matter for VOD?
No — latency is a live-streaming concern. For on-demand content there's no 'live edge' to fall behind, so latency is irrelevant; startup time and buffering matter instead.
Why is live streaming latency higher than broadcast?
Segmented delivery adds it up: encoding and packaging, segment duration, player buffer and CDN propagation each contribute. Standard HLS/DASH stacks accumulate 15–30 seconds, which is why low-latency modes shrink segments and buffers to close the gap.
What's the trade-off of lowering latency?
Smaller buffers mean less protection against network variability, so aggressive low-latency configurations raise rebuffering risk on weaker connections. The right target balances how much latency actually matters for your content against delivery robustness.
Build it with Enveu
Stream live with low latency
Enveu delivers low-latency live at scale with adaptive delivery and multi-CDN.