This feature allows client webcams attached to be exposed to the applications running on the server.
The webcam is not forwarded by default unless the webcam command line option enabled, or the webcam is activated manually from the system tray menu.
Clients only rely on opencv and its python bindings.
The server side is only supported on Linux. It relies on a virtual video device, you must install the v4l2loopback kernel module from and load it:
modprobe v4l2loopback devices=1 exclusive_caps=0
Some distributions may load the module with the wrong setting: exclusive_caps=1
(ie: Ubuntu, see #1596)
The user running the xpra session must be able to access the video devices (ie: usually requires adding the user to the video
group)
The server should work out of the box provided that the kernel module is loaded.
You can enable the webcam from the client's system tray menu, or using the command line option webcam=on
, you can also specify which video device to forward on the command line --webcam=/dev/video2
.
- only supported with Linux servers
- work in progress, see #1030
- framerate is low
- low resolution (hardcoded colorspace dependencies)
- no support for multiple sessions per server..
- limited detection of devices added or removed from the system
- excessive bandwidth usage
- server setup requires an out of tree kernel module
- MS Windows client builds have very unreliable support
Diagnostics
- use the
-d webcam
debug logging flag - run
xpra webcam-info
on the server to locate the virtual video devices:Found 1 virtual video device: /dev/video1
- run
xpra webcam
on the client to run the webcam capture test application (akaWebcam_Test
on MS Windows and MacOS).