Common questions about vcam.
What is vcam?
vcam is a Linux application that enhances your webcam video in real time. It sits between your camera and video conferencing apps (Zoom, Teams, Meet), applying professional-grade image processing to every frame. The enhanced feed is output through a virtual camera device that any application can use.
Which cameras are supported?
Any camera that appears as a V4L2 device on Linux. This includes virtually all USB webcams, built-in laptop cameras, and DroidCam (Android phone as webcam via USB or WiFi).
Can I use my phone as a webcam?
Yes. Install
DroidCam on your Android phone and the DroidCam Linux client on your computer. Connect via USB or WiFi, and vcam will automatically detect your phone as a camera. All 16 processing engines work with the DroidCam feed.
How does the virtual camera work?
vcam uses v4l2loopback, a Linux kernel module that creates virtual camera devices. The processed video is written to this virtual device, which appears as a standard webcam to all other applications. Zoom, Teams, Meet, OBS, and any V4L2-compatible app can select it as a camera source.
Why isn't my camera detected?
Check the following:
1. Ensure the camera is connected and powered on
2. Check if the camera appears as a device: run ls /dev/video* in a terminal
3. Make sure no other application is exclusively using the camera
4. If using DroidCam, ensure the DroidCam client is running before starting vcam
5. Try clicking "Refresh" in vcam's camera picker
How do I enable v4l2loopback?
Install and load the kernel module:
sudo apt install v4l2loopback-dkms v4l2loopback-utils
sudo modprobe v4l2loopback
To make it load automatically on boot, add v4l2loopback to /etc/modules-load.d/v4l2loopback.conf.
Do I need a GPU?
No. All 16 processing engines work on CPU. However, the AI Enhance+ engine benefits significantly from an NVIDIA GPU (GTX 1650 or newer) with CUDA support. Without a GPU, AI Enhance+ runs on CPU at reduced frequency (every 5th frame instead of every 2nd).
Which GPUs are supported?
Currently, NVIDIA GPUs with CUDA support are used for AI Enhance+. AMD GPUs are not yet supported for AI acceleration but all other engines work on any hardware. CPU fallback is always available.
What are the Simple / Normal / Expert modes?
vcam uses progressive disclosure to avoid overwhelming new users:
Simple: Core engines only (Studio Look, Noise Reduction, Mirror)
Normal: Adds White Balance, Tone Map, HDR, Sharpen, Background Blur, Auto Frame, Key Light
Expert: All 16 engines including Color LUT, AI Enhance, AI Enhance+, Eye Contact, Super Resolution
Is vcam free?
vcam is open source under the GPL v3 license. The core application is free. Premium features require a license key, which will be available for purchase when the app reaches stable release.
Does vcam work on macOS or Windows?
Not currently. vcam is Linux-first and uses Linux-specific APIs (V4L2, v4l2loopback). macOS and Windows ports may be considered in the future based on community demand.
Does vcam connect to the internet?
No. vcam never makes unsolicited network connections. The only optional network access is downloading the AI Enhance+ model (~91 MB), which happens only after you explicitly consent to the download.
How do I report a bug?
Please report bugs via the project's issue tracker or contact
zebra@stri.be.