So real quick, Nvidia released an AI tool to take advantage of the ridiculous processing power in their GPUs for some normal everyday stuff. It’s called Nvidia Broadcast and it is actually pretty cool.
It uses AI to enhance the image and audio quality on your computer’s camera and mic as well as apply effects to enhance the audio and video experience. Let me give you a few of the most interesting ones.
Eye Contact
The eye contact feature essentially generates virtual eyes that replace your actual eyes to make it look like you are looking at the camera all the time. You can be looking away from the camera or even have your eyes closed and it will keep your virtual eyeballs glued to the camera.
Auto Frame
They also went a step further and added an equivalent to Apple’s center stage where if you struggle to stay still in a single position, the software uses object tracking to bring you back into the frame. So what it does is it crops into the full field of view of the camera and then moves this cropped section around within the full camera’s field of view to track your face and keep you in the middle of this frame.
One slight drawback is that when you crop into the full frame of the camera’s field of view, you are effectively sacrificing pixels which leaves the overall image less sharp. This is even worse when you realize most laptop webcams and some desktop webcams have very low resolutions of either 720p or 1080p. So if you are using this feature for a live video broadcast like a podcast, you will get a significant hit in video quality unless you have at least a 2K QHD camera.
It also has a video enhancement tool that works on reducing noise in the video especially noise caused by poor lighting conditions.
Echo and background noise suppression
Audio also gets some love. There is echo removal to reduce or eliminate echo from your audio as well as background noise removal which acts as a form of noise canceling for your audience. It will learn your voice pattern and eliminate anything outside of that so you sound a lot more audible.
Stiff minimum requirements
And the amazing thing is this software is processing these audio and video effects in real time which is why the system requirements are pretty high. You need at least an Nvidia RTX2060, RTX Titan, or Quadro RTX 3000 GPU, and as for CPU, at least an 8th Gen Intel Core i5 or an AMD Ryzen 5 2600.
Nvidia broadcast operates as a video and audio source. If you have a preferred camera for video, you can select it as the default camera in Nvidia broadcast. The same goes for your preferred microphone and speaker system. Once this is done, you can open an app of your choice for calls or podcasts and select Nvidia Broadcast for your camera, microphone, and speaker. So your audio and video will be processed by Nvidia Broadcast in the background in real-time and the clean output is what your audience will see. Did I mention that it is completely free? You can download it on Nvidia’s website.
Share
Home
2 comments
ooh great
Aaaaa…sweet sweet Nvidia… I’d kill people to have one of these…I could run my life with that..damn..