Nvidia is improving the suite of features it offers to livestreamers with its GeForce RTX graphics cards. The company introduced a revamped Nvidia Broadcast tool today as part of the rollout of its next-generation RTX 3000-series GPUs. This software uses RTX-accelerated artificial intelligence to help broadcasters improve their audio and video.
Nvidia was already testing some of these features with the recent release of RTX Voice. That beta tool uses AI to remove background noises. Nvidia Broadcast builds that into an improved interface to easily isolate your voice or the voices of your teammates. In a demonstration, Nvidia showed it eliminating the sounds of a blowdryer and a teammate’s keyboard.
But now Nvidia is also bringing in video tools like background-removal and auto-framing. Like with the voice features, RTX GPU’s use their AI Tensor cores to detect the broadcaster and remove the surrounding background. As with Zoom, you can then add your own image or video. Nvidia Broadcast can also track the livestreamer as they move around and keep them in the center of the frame.
These tools are universal, which means you can plug them into any broadcasting software. Nvidia Broadcast should show up as a virtual camera, and you can add that to OBS or Xsplit. You can also even add it to Zoom, Google Hangouts, and other similar video-conferencing apps.
Other tools also have similar capabilities to Nvidia Broadcast. Logitech’s Capture software can track you with auto-framing. Zoom and Snap Camera can remove your background. And Krisp can eliminate your audio. But Nvidia is doing all of that in one app, and it’s often doing it better — at least compared to the tinny and processed sound of Krisp.
Nvidia Broadcast is rolling out soon.
Author: Jeff Grubb
Source: Venturebeat
