RTX Voice is a new plugin currently in beta testing that aims to take advantage of NVIDIA's GeForce and Quadro RTX graphics cards' AI power (presumably through Tensor Cores) to suppress any audio background noise.
 |
Image Source: Alessio Palumbo |
How's it working out? RTX Voice generates an Audio Interactive on your Windows 10 Screen. You may need to point it to your microphone and/or speakers after downloading the software (yes, it can even eliminate background noise from incoming audio, and not just outgoing audio). It's important to remember, however, that you cannot use the RTX Voice Virtual Audio Tool as the default Windows 10 on your computer. This is because it will even filter audio sources that should not be denounced, you will only need to pick the RTX Voice Virtual Audio System for your voice chat applications, such as game audio or music.
The beta supports already
some of the most popular voice chatting and broadcasting applications, like:
OBS
Studio
XSplit
Broadcaster
Twitch
Studio
Google
Chrome
WebEx
Skype
Slack
However, the last four on
that list can have problems when suppressing speaker performance. Indeed,
NVIDIA 's configuration guide mentions that you should actually only enable the
speaker output when it's really needed, as there may be a minor difference in
audio quality in some instances and the RTX Voice app might needlessly take
some resources out of your device.
It is also necessary to
ensure that the built-in noise-canceling effects of your program are turned off
when activating RTX Speech. Notice that you will not be able to monitor the volume of your underlying devices, which means you will need to pick, tweak,
and then pick RTX Voice again to change the volume.
You can also help NVIDIAtrain its AI network by recording yourself for 15 seconds speaking, then recording the environment, and uploading the audio here.
){kind=link}
0 Comments