How to add a Drone
To add a drone or other streaming camera (such as CCTV, bodycams, or IP cameras) to an Eyeson call, you'll need to convert the video stream to WebRTC protocol. Here's how to accomplish this.
Install and execute Ghost
- Download and install the Eyeson Ghost RTMP server from the Github repository
- Launch the Ghost server application
- Find your IP address
Stream from your drone
Configure your drone's streaming application to establish an RTMP connection. The target stream destination requires:
- Protocol: RTMP
- Port: 1935 (default; can be changed with
--rtmp-listen-addr
) - IP Address: The accessible IP address of the Ghost server
- For local network: Use private IP (e.g. 192.168.x.x)
- For external access: Use public IP with appropriate port forwarding (e.g. ngrok)
Note that all streaming devices must have network connectivity to the Ghost server, either through local network access or properly configured public interfaces.
Connect the video stream from ghost to the call
- Start an Eyeson meeting
- Copy the guest link and process it with Eyeson Ghost
rtmp_client --user-id drone01 --delay 40 --rtmp-listen-addr rtmp://`%IP_ADDR%:1935` %GUEST_LINK%
The delay parameter controls stream buffering duration in milliseconds. Higher delay values provide improved stream stability at the cost of latency, particularly beneficial for unreliable network connections. Optimal delay values typically range from 20ms to 200ms depending on network conditions and must be empirically determined for your specific setup.
For all available flags, you can use:
rtmp_client --help
Start streaming
Configure your drone to start RTMP live streaming using your PC's IP address as the target. You can use the user-id to position the stream in the layout.
Remarks on Ghost
This implementation demonstrates adding an H.264 video stream to the call. The current version of Ghost specifically supports H.264 encoding. Since Ghost is open source, developers can extend it to support additional protocols, enabling integration with custom cameras or legacy systems.