Nvidia came out with some exciting news during TwitchCon this weekend. They showcased their next-gen streaming enhancements on their RTX Broadcast Engine. This engine is set to “use the AI capabilities of GeForce RTX GPUs to transform live streams.”
RTX Broadcast Engine
It may be a more apt question to ask what it can’t do! While some of these features may not be groundbreaking for professional streamers, they will certainly help up-and-comers elevate their streams. If they own an RTX card of course…
Nvidia will be able to provide effects in real-time such as greenscreen, augmented-reality avatars, and filters. This is all down to Nvidia’s hardware-accelerated video encoding engine. The best thing about this is that it won’t affect your framerates in-game while using these features.
As mentioned earlier, streamers who have their setup already sorted may not care too much about these features, the main benefit to using this new broadcast engine is that it can be used on a single machine. No longer will you need a separate setup to encode your live stream, this cuts down a lot of the costs for people who are looking at just getting started.
What will I need?
As this is an Nvidia creation you will need to be running one of the GTX graphics cards to use these features as several of the enhancements on offer utilize the Tensor cores. The requirement of the GTX series of cards will be unfortunate for those who aren’t already running them. However, maybe some of the more advanced features Nvidia has to offer will push people to upgrade.
Standout features
One of the features that make this engine stand out is the ability to create your own greenscreen…without a greenscreen! Obviously this cuts down on the costs of actually creating your own green screen, but it also allows you to set this up on the fly (without having to rearrange your office space) with impressive results.
The RTX greenscreen allows for you to “deliver real-time background removal of a webcam feed, so only your face and body show up on the live stream.”. Although (aside from the release promo) we haven’t seen this in action yet, this will hopefully start to convince more people to live stream due to an easier point of entry.
Alongside the RTX greenscreen, we will also be able to make use of the RTX AR feature. This allows the software to not only detect your face but scan for facial features and model the face in real-time. This leads to some pretty impressive real-time augmented reality effects with no extra hardware or software needed. Although this may end up as a mere gimmick for live streamers, the technology itself is quite impressive and may lead to greater AR developments in the future.
Finally, you may also use style filters if you are using these stream enhancements. The RTX style filters allow you “to transform the look and feel of a webcam feed based on the style of another image.” This will allow you to instantly change the look of your webcam stream at the touch of a button. Is this feature essential? No, but it is a cool feature nonetheless.
Existing integration
Nvidia is also currently working with OBS (a popular application for live streams) to integrate some of these features. This is good news for those who are already familiar with this software and don’t want to change too much of their setup.
Now that other companies such as Twitch Studio, Elgato and Discord are starting to use the Nvidia video codec SDK, this is starting to make Nvidia cards very desirable for those whose main system aim is streaming.
Closing thoughts
While some of these features may seem gimmicky on the surface, the technology being used behind them is nothing to sniff at. Personally, I’m excited to see how far Nvidia is going to take this idea and just how well they will start to integrate it into other software.
With their codec being used by other applications such as Discord, who recently released their “Go Live” feature which uses the Nvidia video codec to broadcast games inside the application, this puts Nvidia graphics cards high on the list for budding streamers.
We don’t yet know when these features will become available but we’ll be sure to keep an eye out for any further developments.