Nvidia Shadowplay - How to get 3 Audio Tracks for Recording

https://www.reddit.com/r/letsplay/comments/aofheh/nvidia_shadowplay_how_to_get_3_audio_tracks_for/

Nvidia Shadowplay - How to get 3 Audio Tracks for Recording

Introduction

I have neglected to create this post for a long time, mostly due to my lack of faith in the method. But after a day of troubleshooting, for the purposes of recording gameplay with Nvidia Share/Shadowplay and having the benefit of 3 audio tracks, a feature typically reserved for more professional and performance hungry software, this is the only method by which you can do so until Nvidia adds the option themselves.

If you don’t feel like reading you can just flick through the screenshots and copy the settings shown, but I’m not taking the blame if you miss something important.

DISCLAIMER: I realise OBS + NVENC Encoding is technically better (it has more option + features), this is just for the few random people that would rather use Nvidia Share over OBS.



THE REQUIREMENTS
Voicemeeter Banana/Potato

You're going to need Voicemeeter Banana or Voicemeeter Potato. Either will work, but this guide will be assuming you have Banana installed. The only difference between the two is that Potato has many more options but can be overwhelming if you do not know what you're doing.

With how I’ve set up Voicemeeter in this guide mean that it must ALWAYS BE RUNNING ON YOUR COMPUTER FOR YOU TO HAVE AUDIO. On my i5-6600k this uses anywhere from 1%-5% of my CPU, so if CPU usage is a problem for you abort the mission now.

https://www.vb-audio.com/Voicemeeter/banana.htm

https://www.vb-audio.com/Voicemeeter/potato.htm

Virtual Audio Cable

This program is fundamental to making this work. I recommend downloading VB-Cable and not Hi-Fi, as Hi-Fi in my experience can cause issues with audio (crackling/stuttering).

https://www.vb-audio.com/Cable/index.htm

GeForce Experience

If you’re reading this, I assume you’ve already got this installed. Otherwise, why the heck are you looking up guides for free software you don’t have?

https://www.geforce.com/geforce-experience/download

A Microphone

So we can hear your beautiful voice.

Friends

I’m assuming you have friends, otherwise, why do you need 3 audio tracks?



THE METHOD
STEP 1

Reference Screenshot (Playback Devices) https://pasteboard.co/I0cc1Qm.png

After installing all of the required software, open up your computers sound settings. In your sound settings, there should be 4 tabs; Playback, Recording, Sounds and Communication. Select Playback. This is where your computers sound output devices (speaker/headphone ports) are listed. With the installation of Voicemeeter and Virtual Audio Cable, there should be 3 new devices listed here. These devices are “Cable Input”, “Voicemeeter Aux Input” and “Voicemeeter Input”.

While in this menu, make note of which audio device you’re using to listen to sound on your computer as you’re going to need to know it later.

Right click on Voicemeeter Input and select “Set as Default Device”.

Right click of Voicemeeter Aux Input and select “Set as Default Communication Device”.

Doing this will set Voicemeeter Input to be the audio device that all non-communication programs on your computer will use to play sound. If you can no longer hear anything on your computer, do not worry- this is normal and we will be fixing this later.

Setting Voicemeeter Aux Input to be the Default Communication Device, in theory, makes it so all programs used for communication (i.e Discord, Skype, Teamspeak, etc.) use this device for playing sound. However this is a bit finicky, so you may have to individually set your communication programs to use this sound device.

Once done, close your sound control panel and open Voicemeeter Banana.



STEP 2

To start, a basic explanation of Voicemeeter. Everything explained will be necessary to set up 3 audio tracks and everything not explained is not necessary.

NOTE: Input = Where sound enters or is created, Output = Where sound exits.

Voicemeeter allows for hardware inputs, AKA microphone inputs, to have their sound be directed to specific hardware outputs. Virtual inputs are a form of hardware input, except, as their name implies, they are virtual. There are also virtual outputs that are like virtual inputs but instead output.

A1, A2 and A3 refer to hardware output 1, 2 and 3.

B1 and B2 refer to virtual output 1 and 2.

When A1 is selected underneath a hardware input, all sound from that hardware input will be outputted at hardware output 1. This is the same for all outputs and inputs.

For each hardware input, there is a section labelled “intellipan”. Right click on this setting to cycle through a series of different voice effects, including a voice modulator, right-left stereo panner and a voice colourisation adjuster. If you want to know how these work, play with them in your own time.



STEP 3

Reference Screenshot (Voicemeeter Settings) https://pasteboard.co/I0cd5IW.png

NOTE: I recommend simply copying the settings in this screenshot and skipping to Step 4, but if you want to know exactly what to do please read through the entirety of Step 3.

To begin configuring Voicemeeter, set hardware output 1 (A1) to be your computers main output, aka where sound normally comes out of your computer. In the screenshots, this device is called “Speakers (Realtek High Definition Audio)”. Then set hardware output 2 (A2) as Cable Input.

Next, set any hardware input to be your microphone and then set another to be Cable Output. In the screenshots, my microphone is called “Microphone (2- ATR USB microphone)”.

Underneath Virtual Inputs, click the A1 button on both Voicemeeter VAIO and Voicemeeter AUX. These two virtual inputs correspond to the devices you set as your default and default communication devices in your sound control panel and turning A1 on will allow you to hear the audio from these devices.

You’re also going to want to enable A2 for Voicemeeter AUX.

For both Cable Output and your microphone, right-click on the intellipan menu until you see the panel labelled position. You then should drag the square icon on the panel to the bottom corners for each device. For example, if I drag the square to the bottom left corner of Cable Output’s panel, I would want to drag the square to the bottom right corner of my microphone.’s panel. What this does is force each device to only output to the left or right channel of a stereo sound channel.

Enable A2 and B2 for your microphone and enable B2 for Cable Output.



STEP 4

Reference Screenshot (Nvidia Share Audio Settings) https://pasteboard.co/I0ccA2P.png

Now that the Voicemeeter configuration is over, let’s connect Voicemeeter to Nvidia Share.

Open up the Nvidia Share menu (default shortcut is Alt-Z). Click on the microphone icon and go to settings. Enable “Separate both tracks” and change your microphone source to be “Voicemeeter Aux Output”. Now exit out of Nvidia Share.



STEP 5

Now technically you’re done. But before I get a bunch of confused replies asking how to get 3 audio tracks using this method, I’m going to explain it here and now.

This method does not actually make a 3rd audio track. What it actually does is combine the microphone and discord audio together on one stereo track. Each source gets its own left and right audio channel to itself. To split these tracks up into separate mono tracks (mono audio plays the same sound across all speakers), most video editing software supports the ability to split a stereo track into two mono tracks and if in doubt the free software Audacity has support for this feature.

As far as I can tell, doing this does not decrease the audio quality of your microphone (people with super high-end mics might notice a difference) and the quality of recorded discord chatter is not decreased either. I’ve chalked this up to Discord not actually processing voice in stereo and instead converting to mono (Bots still play sound in stereo) and most mics don’t typically make use of their stereo functionality- and why would you want stereo mic functionality when recording gameplay? (if there is an actual use case let me know).

Regardless, congratulations! You’re done! Enjoy recording using Nvidia Share with 3 Audio Tracks! For those who want a bit more explanation as to how the whole audio system works, read the epilogue found below.



EPILOGUE

With the setup of Voicemeeter, we’ve configured the program to do two things. First if all, it is taking the audio from your computer’s general applications (games, web browsers) and is mixing it with audio from your communication apps (Discord, Skype, Teamspeak, etc.) so that you can hear it through one set of headphones/speakers even though it appears to Windows that they are playing on two completely different devices. In other words, Voicemeeter VAIO + Voicemeeter AUX = Hardware Output.

The second thing that Voicemeet is doing is outputting your communication audio through the Cable Input. Cable Input send it’s audio to Cable Output, which is where the audio is converted to be on a single stereo channel. This allows you to hear the audio in dual channel stereo but have Nvidia Share receive it as only a single channel. The same is done for your microphone, although it is a lot less complex since we don’t have to reprocess the microphones sound, it’s already a hardware input.

These two devices, now on separate stereo channels, are then routed to a virtual output (B2). Your PC treats this as a microphone, so it’s only a matter of selecting this virtual output in Nvidia Share as your microphone and then splitting the audio into mono tracks after recording for editing.