Friday, April 17th 2020
NVIDIA Unveils RTX Voice, AI-based Audio Noise-Cancellation Software
Perhaps the biggest gripe about attending office calls and meetings from home these days is the background noise - everyone's home. NVIDIA developed an interesting new piece of free software that can help those on desktops cut out background noise in the audio, called RTX Voice, released to web as a beta. The app uses AI to filter out background audio noise not just at your end, but also from the audio of others in your meeting as you receive it (they don't need the app running on their end). The app leverages tensor cores, and requires an NVIDIA GeForce RTX 20-series GPU, Windows 10, and GeForce drivers R410 or later. RTX Voice runs in conjunction with your meetings software. Among the supported ones are Cisco Webex, Zoom, Skype, Twitch, XSplit, OBS, Discord, and Slack. For more information and FAQs, visit the download link.
DOWNLOAD: NVIDIA RTX Voice beta
DOWNLOAD: NVIDIA RTX Voice beta
57 Comments on NVIDIA Unveils RTX Voice, AI-based Audio Noise-Cancellation Software
Some TPU members: Free? What is the catch? Are you transmitting CRISPR protein to my brain via your evil 5G based signal to give me testacle cancer and look at my porn browsing history??
Also, since I'm not into electronics, I have no idea what the parts you've mentioned do. So another question: you expect me to learn that and build a preamp? Or what exactly?
Why do you think utilizing AI is harder to implement than building some circuitry is beyond me.
Part of me just don't believe that Nvidia is stupid enough to make a spyware and hide it, when there will be a backlash for doing so. With GPP they already know that people aren't going to get behind every bulls**t that they do.
As far as I'm concerned it's always nice to be skeptical about big govt & megacorps, there's about an even chances that your fears aren't dumbfounded based on the other oft-used phrase ~ Just Because You’re Paranoid Doesn’t Mean They’re Not Out to Get You :toast:
If you look into Snapdragon HEXAGON includes custom DSP dedicated to the task, some things taken from Conexant IP. If you remember the HP scandal with Superfish, a keylogger that came in... it resided exactly in the MIC recorder app from Conexant audio driver package and started on such low ring level... Well... khrmm... like it didn't happen before eh?
The most gentle way is to press a damn button on the device - MIC... and everything works... There is no software layer needed, no drivers, no latency, zero issues. RTX? AI? Quit joking people.
And FFS it is snowing outside...
For example there's a nuclear solution like Cirrus Logic CS47L85 includes HW algos even for word detection, various filtering etc etc, no RTX, no AI. Actually Apple uses the Cirrus Logic IP and they know what they do as Industry vets.
But other than that... we have usually have it included already... even in the Realtek IC. But the point is the ADC part is made so poorly often, look at mobo reviews, the level is like early 70ties in SNR department, nobody understands the numbers, the implementation, you cannot turn dirt into gold. The ALC's do support all the DSP techniques also. But their drivers... and on top of it another layer...? Windows is already full of rubbish.
Why on earth other consumer device have made proper solutions that work fine? There shouldn't be issues in the first place for nvidia even to bother about it.
It is a shame your other points are right and true , it's a shame so much tech gets poorly used.
Have fun!
So extreme it doesn't concern you unless you are a drug baron, alt right extremist or terrorist, but dammmmnn still.
Also, you're throwing more and more information about noise cancelling hardware and so on. About buying better microphones and preamps. Seriously. Do you expect people to get all that stuff for work?
I'm sure many gamers decide to burn some more money on streaming equipment after they fill saturated with fps, but what Nvidia skunkworks provided works with any microphone. That's its second biggest strength. The first one is what it's actually doing and what you don't understand. :)
I bet this guy doesnot even use VPN 24/7 with kill switch.
I have a laptop, I use it for meeings. Tell me how I can make all the unwanted sound disappear without software and "skunk work". :)
And once again, whatever the hardware you suggested does, it surely doesn't replicate what can be done with software. Not possible.
You can also run the app offline, in case anyone was wondering. It looks like its going through the Core Audio API in Windows 10.
Muting your own output works really stupidly, it doesn't tidy up your mic output it muted your whole PC's sound output, kinda useless.
But it was configurable to only work on incoming sound, like this it works well, though I turned it down half way to reduce vocal cutouts.
This was using discord.