Time Synchronization

Multi-display playback depends on all nodes sharing an aligned clock. Without it, content on different Runner nodes drifts apart, and even a small offset shows as misalignment across adjacent displays.

WATCHOUT uses NTP (Network Time Protocol) to align the system clocks of all nodes.

Synchronization modes

  • WATCHOUT (default) — WATCHOUT configures NTP automatically. The node running the Director becomes the NTP server (the time reference), and every other node syncs to it as a client. Simplest and recommended for most installations.
  • User (advanced) — you manage NTP outside WATCHOUT. WATCHOUT does not change the node's NTP settings, and the NTP Server field stays blank. Use this when the nodes belong to a larger infrastructure with its own time policy.

In User (advanced) mode, an incorrect external NTP setup makes playback drift. Use it only if you are confident in your NTP configuration.

Setting the sync mode

  1. Open the Nodes window and select one or more nodes.
  2. In Node Properties, open Sync Settings.
  3. Choose WATCHOUT or User (advanced).

Changing the mode may restart services on the affected nodes.

NTP server

In WATCHOUT mode, the NTP Server field (Node Properties → NTP) shows the server a node syncs to, and WATCHOUT manages it. To make the show follow an external clock, point the time-source node at an upstream NTP server here — for example a facility clock. In User (advanced) mode the field is blank, because WATCHOUT does not manage NTP.

The time source

In WATCHOUT mode, the node running the Director is the time source. Its NTP offset shows "N/A (is WATCHOUT time source)" instead of a number, because every other node is measured against it.

When a node stops being a managed NTP client — you switch it to User mode, or it is removed from the synced set — WATCHOUT restores the NTP server it used before. If no previous server was saved, it falls back to time.windows.com.

Monitoring the NTP offset

Each node shows an NTP offset (the difference between its clock and the reference) in the Nodes window and in the NTP section of Node Properties. Under normal operation it is usually within a millisecond. A large offset means a sync problem.

Automatic resync

When a node's offset exceeds about 100 ms, WATCHOUT triggers a resync to bring the clock back in line. It logs the error "NTP offset is too high (Ns) - resyncing.", where N is the measured offset in seconds. This can happen after a node has been offline, after a network interruption, or when the system clock drifts.

Hardware sync

For LED walls and edge-blended arrays, NTP alignment alone is not enough. NTP aligns each node's system clock to within about a millisecond, but it does not align GPU output. Each node's GPU scans out on its own clock. The outputs between nodes therefore sit in different phases — a fraction of a frame apart, even when every node uses the same GPU, driver, and Windows version. Different Windows versions and GPU vendors also implement Full Screen Optimizations (FSO) differently, which changes GPU latency and widens the gap.

A sub-frame phase offset is small, but it shows as tearing across adjacent displays. To remove it, frame-lock the nodes with a Hardware Sync Group. This needs an NVIDIA sync board on each node, set up in NVIDIA's tools first. Create and assign sync groups in Show Properties → Hardware Sync Groups.

A separate Genlock SDI setting locks SDI output timing — see SDI Output.

Some DisplayPort-to-HDMI converters do not handle the synchronized signal well and reintroduce tearing or noise. See Tearing or Noise with Hardware Sync.

PTP and ST 2110

NTP keeps node system clocks aligned for playback. ST 2110 video uses a separate, more precise clock — PTP (Precision Time Protocol) — that runs only between ST 2110 senders and receivers. PTP does not replace NTP. Configure PTP in the ST 2110 Interfaces dialog. See Setting Up ST 2110.

Previous Next