If you're into streaming, whether as a viewer or as a creator, you'll no doubt be familiar with a green screen. Streamers will sit in front of one and by then applying a chroma key in software to the color of the screen, you can make the background disappear and all that's left is you.
The trouble with a green screen, though, even ones as convenient as the Elgato Green Screen, is that they're often expensive and take up a fair bit of space. There have been alternatives using software before, but rarely have they ever been worth any serious time.
XSplit is hoping to change all that with VCam. It's a virtual green screen and background replacement application, initially available to XSplit Premium subscribers. But the best part is that you don't need XSplit to make use of it. That and it's actually fairly good.
What VCam does and how you get going
Much like other attempts at this, VCam operates as a separate application that can then be seen as a webcam in your other software. For example, in OBS, you wouldn't add your actual webcam, you'd add VCam as the source instead. The only way it works is if the VCam app is acting as a go-between.
Setup is very simple. Right now you need to be an XSplit Premium member since it's in its early stages, so you'll need to login using whatever your registered XSplit account is. The app will take you through a short setup process where it checks your PC to make sure you can run it OK.
This is a fairly resource intensive app, as such you'll need at least a 6th generation Intel CPU, preferably an i5 or i7, Windows 10 64-bit and a fairly recent GPU, from NVIDIA's 8-series and up. I've tested it with a Ryzen 7 2700X and an NVIDIA 1080 Ti, but whatever hardware you're on you get the chance to scale the quality so you're not maxing out your PC.
A great start with room for improvement
As far as the results go, it's pretty impressive. Certainly better than the Logitech effort that comes with the C922 webcam, though lighting is obviously key. Using my Razer Kiyo to give me a healthy orange glow, VCam didn't seem to struggle to detect any edges.
Whether blurring, replacing or removing the background, I'd have to give it a thumbs up. It isn't perfect, though honestly, we're still a way off software being as good as an actual green screen. But this is more than good enough to be worth thinking about using.
The main issues seem to come, unsurprisingly, at the edges. Detection is great, but it's where I noticed a lot of flickering on my white headset. If I take that off, it seems to be much less noticeable, but that's the headset I wear when I'm gaming so it's certainly something I'd like to see improved upon.
I'd also like XSplit to be able to reduce the amount of CPU power the app uses. On my eight-core Ryzen 7 2700X the higher quality settings could spike to over 60% CPU usage, with the standard setting resulting in a less thirsty 30-35%. While I have headroom to spare, plenty of others who might want to use this on their streaming PC will not.
The good news is that using the GPU produces no lesser quality in the end result and never went above 4% usage on my GTX 1080 Ti, so this is definitely the way to run it in my opinion.
Try it out now
I'm impressed with VCam, and genuinely think it's the best piece of software like this I've tried. It works with XSplit, Skype, OBS, Discord and likely many other applications, and it's still in its early beta stages so can only improve from here.
If you're an XSplit Premium subscriber you can hit the link below and try it out for yourself right now.
Richard Devine is an Editor at Windows Central. A former Project Manager and long-term tech addict, he joined Mobile Nations in 2011 and has been found on Android Central and iMore as well as Windows Central. Currently you'll find him covering all manner of PC hardware and gaming, and you can follow him on Twitter and Instagram.
Will the Kinect camera make it work better?
Get the best of Windows Central in in your inbox, every day!
Thank you for signing up to Windows Central. You will receive a verification email shortly.
There was a problem. Please refresh the page and try again.