[ad_1]
NVIDIA announced today, at TwitchCon, exciting new technologies that could be very useful to the live broadcast community. The company is exploiting the RT and Tensor cores of its current generation of Turing-based GeForce RTX graphics cards to enable features such as real-time raytracing, but this same hardware will be used to power its new RTX Broadcast Engine.
RTX Broadcast Engine will use built-in Tensor cores for a variety of real-time effects, including style filters, augmented reality and green screens. In the case of RTX Greenscreen, NVIDIA uses artificial intelligence to clean up the background of the live stream, leaving only your face and body to overlap with game content or any other background of your choice. The main benefit of using GeForce RTX to enable virtual green screen functionality is that no additional hardware is required for configuration and maintenance, which reduces the cost of input for streamers seeking to improve their channel.
Of course, the greenscreen effect seems almost perfect in the above NVIDIA demo, but we will have to wait to see how everything will unfold in the real world once the software is updated to take advantage of the SDK.
RTX AR uses face tracking technology on your webcam sequences to transfer your facial movements (lips, eyes, etc.) to a 3D model. Think Apple Memojis and you have a general idea of what to expect. Finally, RTX Style can transform the appearance of your feed based on the filter you have chosen. This sounds a little gimmicky, but it might appeal to a live audience segment.
NVIDIA said 750 million people around the world were using their streaming platform to watch the world's best players compete against the competition. The RTX broadcast engine definitely has potential, but application developers will have to add support in their software. In the case of RTX Greenscreen, support is already planned for the Twitch-centered OBS. StreamLabs and XSplit have also integrated support for RTX Broadcast Engine and associated SDKs.
[ad_2]
Source link