How to use NVIDIA RTX Voice to suppress keyboard and background noise

Dear reader, Borderpolar has gone streaming and YouTubing, and if you are doing the same thing, you probably want to tackle the challenge of suppressing background noise and your keyboard taps and mouse clicks while recording or streaming. Thankfully, there is an easy solution to this problem if you have an NVIDIA graphics card, which is more often than not the case for gamers and streamers.

Table Of Contents

NVIDIA RTX Voice installation and setup

Here are the steps to using RTX Voice for noise suppression:

  • First head to NVIDIA's RTX Voice setup page. There you will have two options:
    • If you have an NVIDIA RTX GPU or higher you will need to download NVIDIA Broadcast and I will make another guide for people using NVIDIA Broadcast.
    • If you are like me and have an NVIDIA GTX GPU – I have an NVIDIA GTX 1660 Ti – you will have to download RTX Voice.
  • Install RTX Voice.
  • Configure RTX Voice. In RTX Voice's configuration, you will have to select the audio input device and enable the “Remove background noise from my microphone” option.
RTX Voice configuration
  • In the image above I have elected to suppress background from my webcam's microphone.
  • In case you are streaming with someone else and you want to suppress the background noise coming from their source you may need to enable the second option to suppress noise from your speakers or headphones, but it is highly recommended that the other person does instead. This will be less straining for your GPU.

RTX Voice uses AI to remove background noise. Essentially, RTX Voice is a software that acts as an intermediary between your camera and your streaming software.

So the next step is to add RTX Voice as an Audio Input Device to your streaming software and in this guide I will show you how to do it for OBS Studio and Streamlabs OBS.

Personally, I use Streamlabs OBS as I had issues with streaming and recording quality in OBS Studio that were instantly resolved by moving to Streamlabs.

Adding as RTX Voice as an Audio Input Device (OBS Studio)

To add RTX Voice as an Audio Input Device you will need to start OBS Studio first.

  • Next, go to the “Sources” window at the bottom and click the + icon to add a new source.
  • In the popup menu, select “Audio Input Capture”.
  • In the Create/Select Source window, choose “Create New” and add a name to the source e.g., RTX Voice and click “OK”.
OBS Studio – Select/Create Source
  • In the next window choose “Microphone (NVIDIA RTX Voice) from the dropdown menu.
OBS Studio – Select NVIDIA RTX Voice as Audio Input Capture Device

  • Click “OK,” and the “RTX Voice Tutorial” (or the name you chose) will appear in your “Sources” window. We are almost done.
  • There is one problem, though. The software processing performed by RTX Voice creates some delay. If you record your voice without taking into account that delay, your voice will sound distorted as it is still in the middle of processing. Don't worry; the solution is simple.
  • Go to OBS Studio's audio mixer, click the cog of your RTX Voice audio input capturing device and select “Advanced Audio Properties.”
OBS Studio – Select “Advanced Audio Properties” for your RTX Voice audio device
  • What we want to do here is to add a minor sync offset value to make up for the delay caused by NVIDIA RTX Voice's processing. I have found “100ms” to be a sweet spot, but you may have to go up to “200ms” if your graphics card is slow or the background noise too loud. Don't go too high, though. Otherwise, the video and audio won't be in sync anymore.
OBS Studio – Set the Sync Offset for your RTX Voice device to 100ms.
  • You are ready to start recoding or streaming!

Here is my video guide for the process. Notice the distortion in my voice due to the extremely loud background noise at the start of the video (I am recording on a laptop without an RTX card so RTX Voice could not do an excellent job with that sound level).

NVIDIA RTX Voice setup for Streamlabs OBS

The process for Stremlabs OBS is almost identical to OBS Studio. Here are the steps to add RTX Voice and add a sync offset to your audio capturing device:

  • Start Streamlabs OBS
  • Go to sources and press “+” to add a new source.
  • In the “Add Source” window selected “Audio Input Capture” from the “Standard” list.
Streamlabs OBS – Add Audio Input Capture
  • Toggle the option to add a new source and give a name to your source.
Streamlabs OBS – Add New Source and give it a name
  • In the next window you will need to select “Microphone (NVIDIA RTX Voice). Then Click “Done” and your source will be added.
Streamlabs OBS – Select Microphone (NVIDIA RTX Voice)
  • Now we need to apply the 100ms sync offset to “RTX Voice Tutorial.” Go to the “Mixer” window and press the cog at the top right corner with the snippet “Open advanced audio settings.”
Streamlabs OBS – Open advanced audio settings
  • In the pop-up window set the Sync Offset for the device to 100ms.
Streamlabs OBS – Advanced Audio Settings. Set “Sync Offset” to 100ms.
  • That's it! You are good to go!
Cross the borderline!
CURRENTLY OFFLINE
Exit mobile version