Skip to main content

Add your first camera

Add cameras gradually. A single working camera is more useful than many half-configured cameras.

Start with one camera, validate snapshot and stream access, publish the source for live viewing, then place and map it in the composition. Build event pipelines only after live viewing is stable.

If you do not have a real camera ready, use Test without a real camera first. It gives you a synthetic RTSP source that exercises the same manual camera and live viewing flow without adding test-only behavior to the product.

Before you start​

Have this information available when possible:

  • camera name;
  • IP address or hostname;
  • RTSP URL, if you already know it;
  • ONVIF username and password, if the camera supports ONVIF;
  • whether the camera supports PTZ;
  • where the camera is physically installed.

Use a contained network and avoid exposing camera credentials publicly.

Discovery or manual setup​

Toposync can work with cameras added manually and, when supported by the camera and network, ONVIF discovery.

Use discovery when it works, but do not depend on it for every camera. Some devices, VLANs, Wi-Fi networks, routers, or camera settings can block discovery even when the camera itself works.

If discovery does not find the camera, add it manually with the information you already have.

First validation​

Before building pipelines, confirm the basics:

  1. The camera is saved with a clear name.
  2. Snapshot or image access works.
  3. At least one video source works if you installed streaming support.
  4. The source has the right role: main, sub, zoom, or custom.
  5. Transmit this source is enabled for the sources you want to view.
  6. The live view appears as live, warming up, or with a clear actionable problem.
  7. PTZ controls work only if the camera supports PTZ and credentials allow it.
  8. The camera is placed near its real location in the composition.

Do not tune advanced settings until this first loop works.

Publish the source for live viewing​

For normal camera viewing, do not create a technical transmission manually. Publish the camera source and let Toposync generate the transmission and continuous implicit pipeline.

Common first setup:

  • low-resolution stream as sub for grid tiles;
  • high-resolution stream as main for fullscreen;
  • PTZ/autotrack stream as zoom when the camera supports it.

Open the dashboard and confirm that the camera is visible. A short warmup is normal; a persistent No pipeline or Stale state needs attention.

See Live viewing for playback states and transports.

Place the camera in the composition​

Placement does not need to be exact on the first day. It should be good enough to answer:

  • where is this camera installed?
  • what area does it mainly see?
  • which rooms, gates, doors, or outdoor zones are visible?
  • which events should belong to this camera?

After the camera is placed, revisit your areas. You may need to split or rename areas based on what the camera actually sees.

Map the camera when you need spatial context​

Mapping is optional for basic live viewing, but required for 360 View and for pipelines that use camera-to-world positions.

Map the camera after the source and live view are stable. Start with four corner points, then use internal refinement points only when the image needs local correction.

See Camera mapping for calibration and 360 View behavior.

Build pipelines later​

After live viewing works, create pipelines for events and automation:

  • motion or object detection;
  • area filters;
  • stored images or crops;
  • notifications;
  • Home Assistant actions;
  • custom rendered live variants through Transmit video.

Use Create your first pipelines for the next step.