How-to: Record desktop and Discord/Skype sounds in separate audio tracks for video editing

Cassius Sonoda

by Cassius Sonoda

YouTuber, blogger, PC hardware seller, and PC gamer.

Problem scenario

You are creating videos for an internet audience, and you have identified a need to be able to treat sound different sources individually: your voice in the video, your voIP partner’s voice, and sounds coming from the desktop.

For example, there are parts in the video where your voice is getting drowned out by the in-game/app music/sounds, or your friend is talking over you on Discord.  Imagine having the power to amplify your voice and lower theirs.  Maybe you want to play the video to popular music instead, and it is interacting badly with the in-game sounds, which you would in hindsight prefer to mute.  Well, it turns out you can.  From now on, you will have much more flexibility in editing by recording separate sound sources as separate audio tracks that will be stored in a singular MP4 file along with the visuals.  Here’s how: 

Apps and app accessories required for this tutorial

The following OS and software are presumed installed:

Summary

  • You will be working with an OBS-generated MP4 video file containing multiple audio tracks
  • Audacity is used to extract the multiple audio tracks from the MP4 video, and output them as MP3 files
  • You can then import those MP3 files into a video editor which such as Camtasia Studio.  Note that more upmarket video editors (e.g. Sony Vegas) may be able to natively pull out the audio tracks without the need for the intermediary step using Audacity.  So if you are fortunate to be using upmarket video editors, you would not need to worry about the extracting the audio step.

Setting up audio devices in Windows

R-click Speakers icon in system tray > Playback devices

Ensure the following devices are enabled:

  1. your currently-used headphone or speaker
  2. VB-Audio Virtual Cable
  3. VB-Audio VoiceMeeter

Ensure (3) is set as Default Device
Ensure (1) is set as Default Communication Device

R-click Speakers icon in system tray > Recording devices

Ensure your recording mic is enabled and set as Default


Setting up your comms apps to output to VB-Audio Virtual Cable (instead of outputting to where all the other desktop sounds are going)

For Discord:

User Settings > Voice & Video
Set output device to VB-Audio Virtual Cable

For Skype:

Tools > Options > Audio Settings
Set speakers to VB-Audio Virtual Cable


Setting up Voicemeeter

Open Voicemeeter.  Note it is important to leave Voicemeeter minimised and running on your desktop all the time for its effects to be active.

The settings are:

Set Hardware Input 1 to your recording microphone

Set Hardware Input 2 to VB-Audio Virtual Cable; and
set “A” active

For Virtual Input, set “A” to active

For Hardware Out, set A1 to your currently-used headphone or speaker


Setting OBS up to record separate audio tracks

In OBS:

Under mixer pane, you will need to assign your audio devices as follows.

Screenshot of OBS context menu where to assign each of your audio devices.
Click to enlarge.

Mixer pane > Desktop Audio > Mechanical cog icon> Properties > select Device: Speakers (VB-audio Voicemeeter VAIO).  This will be all desktop-associated sounds, minus Skype/Discord output, which will be re-routed to VB-Audio Virtual Cable.

Mixer pane > Mic/Aux > Mechanical cog icon > Properties > select Device: Speakers (your recording mic)

Sources pane > Add Audio Output Capture (“+ button, Add”), name: “VB-Audio Virtual Cable” > Properties > select Device: Speakers (VB-Audio Virtual Cable)


Next, to switch on tracks:

Settings (important: ensure Output Mode is set to advanced or the Streaming/Recording/Audio/Replay Buffer tabs will not show) > Output > Recording
Check audio tracks 1- 4

Aside: under Settings > Output > Audio, you can set the desired audio bitrate and name your tracks.


Next, to assign your tracks:

Then click any of the cogs for any of the audio sources under mixer > Advanced Audio Properties and check track boxes to the following effect:

Settings > Output > (Ensure Output Mode is set to “Advanced”) > Recording
Set recording format to MP4
As per the warning, not all video formats will support multiple tracks, and this tutorial presumes you will record videos in MP4 format.

You are now ready to record.  After you stop recording, the output will be an MP4 file with 4 audio tracks as defined above.


Extracting the audio tracks from the MP4

To extract the audio tracks into individual MP3 files, you will use Audacity with the relevant plugins:
Just download (http://lame.buanzo.org/) and install the LAME MP3 and ffmpeg plugins for Audacity.  Fairly straightforward, but should you encounter any troubles see official Audacity instructions:
LAME MP3 – http://manual.audacityteam.org/man/installing_and_updating_audacity_on_windows.html#winlame
ffmpeg – http://manual.audacityteam.org/man/installing_and_updating_audacity_on_windows.html#winff


Having installed the relevant plugins, either open the MP4 file (the one created from your OBS recording session above), or drag the MP4 file into empty Audacity instance.  You will be prompted to select the tracks to import.  Select all the tracks or just the ones of interest if you know which ones they are, and hit OK.

Output will be similar to this:

Note: if you want to test-play or export the tracks at this point, do it solo for each track, since, if there is any overlap between the tracks, for example, more than one of the tracks contains your voice in it, then you may experience a weird-sounding audio artifact, if you play those tracks at the same time.

To export multiple tracks at once, use File > Export Multiple:


Handling the audio tracks in Camtasia Studio 9

Now import the resulting MP3 tracks into Camtasia.  You can disable the sound in the original MP4 file which will only be used for footage, and use the separate audio MP3 tracks you just created.  You should ensure all the MP3s and the MP4 are aligned initially so they are in sync.  Edit as needed.


I’ve helped you – now a small way to return the favour…

I hope this tutorial was helpful.  Please consider subscribing to my YouTube channel if you found it helpful, as a way of thanking me.  It costs nothing to subscribe and is all I ask in return.  Although I don’t typically do technical tutorials on the channel; I do a few from time to time, so it’s worth it.  The channel is more broadly focused on PC gaming.

You can also support me by buying your PC hardware from my store, which offers decent prices on everything computer-related that you could need.


Unresolved problem for geniuses reading this…

As of this time, I still don’t know how separate in-game sound from in-game music.  If anybody has any idea how to do this, please educate me using our contact form.∎

Leave a Reply

Your email address will not be published. Required fields are marked *