When Nvidia launched the free Nvidia Broadcast app in 2020, it promised to make use of the AI capabilities of RTX GPUs to spice up the video and sound high quality of customers’ recordings, presuming they’d the fitting {hardware}. This might be helpful when a laptop computer’s built-in webcam and microphones could not seize the most effective wanting footage or the richest audio, and will probably assist streamers get by with a less expensive setup. With model 2.0 of the Nvidia Broadcast app, launched on the tail finish of January, the capabilities are stepping up even larger…maybe just a little too near the solar.
In Nvidia Broadcast 2.0, microphones nonetheless have choices for noise removing and room echo removing, however there’s now additionally a studio voice impact in beta that “enhances the standard of your mic to simulate a excessive finish recording studio.” For video, the app can nonetheless make tweaks to your background (changing, blurring, or primarily green-screening it), take away noise from grainy footage, zoom in and routinely hold you in body, and surprisingly sufficient, make it appear like your eyes are wanting on the digital camera. However new with the replace is digital lighting, to higher spotlight you in your video.
A few of these new options name for highly effective GPUs. Studio voice, video noise removing, and digital key mild all name for a “high-end GPU” and aren’t really helpful to be used whereas gaming or utilizing different GPU-intensive functions. All options require RTX {hardware}, that means you’ll want at the least an RTX 2060 or above to strive them, however for the options that require a “high-end GPU,” Nvidia says an RTX 4080, 5080, or larger is “required.” That mentioned, I used to be in a position to run each options on an RTX 4060 cellular GPU.
Nvidia Broadcast in motion
The thought behind these AI options is cool, however how effectively they really work remains to be in query. For one factor, they could actually be as demanding as Nvidia says. Working both the digital key mild or studio voice function on my RTX 4060-powered laptop computer confirmed the GPU was completely slammed by the method. Nvidia’s built-in GPU Utilization monitor was exhibiting crimson, with the RTX 4060 all however maxed out and the efficiency overlay exhibiting it drawing 60 watts. My laptop computer’s followers even ramped up as if I have been gaming at full throttle. So simply from an economics standpoint, these options are expensive irrespective of the way you take a look at them. You’ll must have highly effective {hardware} to run them, after which run that {hardware} arduous. Plan on utilizing these options on a desktop laptop or together with your laptop computer plugged in.
Then there’s the much more essential matter of how they actually look and sound. Let’s begin with video.

Credit score: Mark Knapp
The attention contact software, regardless of being obtainable earlier than Broadcast 2.0, has now come out of beta. However I’m not satisfied it ought to have. Certain, enabling it makes it appear like I’m staring into the digital camera in video footage. However in my testing, it invariably gave me massive blue eyes that made me appear like a White Walker proper out of Sport of Thrones. For reference, I do not need blue eyes. Even once I was making eye contact with the digital camera, Nvidia Broadcast nonetheless insisted on modifying my eyes and making them blue.

Credit score: Mark Knapp
The digital key mild did what it mentioned. It created synthetic lighting to spice up brightness on me with out bumping up the brightness on the entire video. The outcomes did not impress me, although. With it enabled, I merely appear like I’ve gone radioactive. The lighting may be very unnatural.
As for the audio, at first blush, it sounds pretty spectacular. The mics on my laptop computer are usually not excellent. Even in a quiet room, they put out audio that has me sounding distant and barely muffled. With studio voice enabled, my voice finally ends up a lot fuller and clearer sounding. However listening intently, there’s an odd digitization occurring. It’s arduous to characterize, nevertheless it doesn’t sound prefer it’s actually my voice. It sounds extra like a recording of my voice was utilized in a voice cloner, after which that repeated all the things I mentioned. It’s all just a bit stilted and quavering. Pay attention beneath:
The studio voice function can also’t save the mic from a foul recording setting. Testing in a small room with a field fan working at full blast, the audio was a dramatic enchancment in readability in comparison with the uncooked recording from the microphones, nevertheless it was nonetheless audibly processed, and the efforts to subdue the background noise made my voice sound particularly odd.
If in case you have a half-decent microphone, studio voice would possibly even make it worse. I made further take a look at recordings utilizing the built-in increase mic on the Audeze Maxwell headset with it straight wired into my laptop computer. In each a quiet and loud room, it supplied a loud, clear, and full recording of my voice with out studio voice enabled. In each instances, turning on studio voice then launched hard-to-miss digitization that not solely made the audio sound worse but in addition made it more durable to grasp.
Can Nvidia AI change a correct streaming setup?
Given the {hardware} necessities, efficiency calls for, and high quality of the outcomes, the celebs actually need to align for these newer Nvidia Broadcast options to really feel actually worthwhile. If in case you have an Nvidia-powered system, by all means, play with the software. Among the options can turn out to be useful, just like the auto-framing one. However I wouldn’t suggest shelling out for a brand new Nvidia GPU simply so it can save you cash on audio and video recording gear, particularly if you wish to get wherever near skilled high quality. And don’t overlook that the facility draw of the GPU making an attempt to run these options will add up over time.
The audio high quality I obtained from studio voice — maybe restricted by the RTX 4060 in my system — wasn’t one thing I’d need to share with any sort of viewers frequently, and it paled compared to the standard I might get simply from having a headset with a increase mic. I’ve examined so much of gaming headsets, and even very low cost wired headsets with a increase mic are leagues higher than what I heard from studio voice.
The attention contact function did not be something aside from unsettling, and I don’t suppose it’s going to idiot anybody into believing you’re truly making eye contact with them. And the digital key mild, very like studio voice, doesn’t seem like a high quality substitute for an actual key mild, particularly when reasonably priced LED lights are a dime a dozen.