Skip to content
Add a Live Stream

Add a Live Stream

A live stream is one channel: a name, one or more sources, and the servers that deliver it. Adding one takes a minute; this page also covers the options that make a channel resilient — backup sources, automatic restarts and on-demand start.

Open Live Streams → Add Stream.

First, one decision: managed or direct

Everything else on the screen follows from this choice.

Managed (the default) — the server ingests the source and produces the outputs your subscribers play. You get failover between sources, on-demand start, TV archive, transcoding, recording and monitoring.

Direct source — the panel hands the player the source URL and steps aside. Nothing is processed, so nearly every advanced option is disabled and greyed out: no archive, no transcoding, no FPS monitoring. Use it for sources that are already delivery-ready. A separate direct proxy switch relays the traffic through your server instead — the source stays hidden from the player, at the cost of your bandwidth.

Details

Name the channel, set its icon (a URL or an upload), and place it in categories and bouquets — both fields create new entries on the fly if you type a name that does not exist yet. Notes are internal, never shown to subscribers.

Sources

Add source URLs in the order you want them tried. The list is drag-to-reorder, and the order is the failover order: if the first source drops, the stream restarts on the next one. A stream with one source is a stream with no plan B — add a backup where you have one.

Instead of URLs, a stream can also capture from hardware — pick a server and its video and audio capture devices.

The advanced options that earn their keep

The Advanced tab is long; these are the ones that matter most in practice.

Reliability. Schedule an automatic restart on chosen days at a chosen time — useful for sources that degrade over a long run. Restart on FPS drop watches the frame rate and bounces the stream when it falls below your threshold, which catches the frozen-picture failure a plain up/down check misses.

Timeshift. A delay in minutes turns the channel into a delayed copy of itself — pair it with a second stream of the same source to offer a +1 channel.

Identity. Per-stream user agent, HTTP proxy, cookie and custom headers, for sources that are particular about who connects.

Adaptive link. Attach another stream as an alternative quality; players that speak adaptive delivery can switch between them.

Title sync. Mirror the programme title from another stream’s EPG — for regional variants of the same channel.

Transcoding. Pick a transcode profile to re-encode rather than pass through. Leave it off where you can: an untouched channel costs no CPU.

Custom FFmpeg. Full control of the command line for the cases the form does not cover. The panel stops validating what it cannot see — test on one channel before rolling out.

EPG

Bind the channel to a guide: pick the EPG source, the channel ID inside it, the language, and an offset in hours if the guide’s timezone differs from the feed. The listing then follows the channel everywhere — panel, playlists, player apps.

RTMP push

Push the channel out to external RTMP destinations as it runs — one or more target URLs, each with its own key. This is an output, independent of what subscribers receive.

Servers and delivery

The server tree decides where the channel runs. Drag servers between enabled and offline; the channel starts on every enabled one.

On demand — instead of running around the clock, the channel starts when the first viewer asks for it and stops when the last one leaves. Idle channels cost nothing; the first viewer waits a few seconds. For faster on-demand starts, LLOD v2 skips source probing and hands the first source straight to the encoder — the trade-off is that if that source is down, the stream does not start.

TV archive — choose which server records the channel and for how many days. Recording runs on the archive server you pick, so a busy delivery server is not also the recorder.

Thumbnails — which server generates the live preview images.

Import many at once

Add Stream → Import takes an M3U and creates streams in bulk; everything lands in the review queue first, where categories, bouquets and EPG mapping are confirmed before going live.

After saving

The stream appears in Live Streams, where the status column tells you what each channel is doing — online, starting, down, on demand, direct, stopped — and the row menu starts, stops or restarts it. Click through to the stream view for uptime, per-server status, bitrate, connected viewers and errors.

A channel is only as good as its guide, so wire up your EPG sources next — and when one server is no longer enough, load balancers install from inside the panel. Building tools instead of channels? The API drives everything on this page programmatically.