← All posts

TikTok LIVE audio mixer: balance mic, game and music

For TikTok LIVE on a PC, the audio mixer you need is already on your machine. You set it up in three layers, in this order: each app’s volume in Windows, one fader per sound in your streaming app, then ducking so the music drops when you talk. After that, you check it by ear.

This post covers the PC side. Going live from a phone with a hardware mixer is a different setup, and it isn’t covered here.

1. Set each app’s volume in Windows first

Open Start > Settings > System > Sound and select Volume mixer. Under Apps, every program that’s playing sound has its own slider and its own Output device.

Check three things for your game, your music player and your browser. The app isn’t muted, its slider is up, and its output device is the one you expect. Microsoft gives this same check as the fix when one app has no sound, so it’s the first place to look when a source goes quiet.

Do this before you touch your streaming app. If a game sits at 20% in Windows and 100% in OBS, you’ll be adjusting the same volume in two places.

2. Give every sound its own fader

Out of the box, OBS captures your desktop audio and your mic. Desktop Audio is one fader for everything: the game, the music, Discord, the ping from a browser tab. You can’t turn the music down without turning the game down too.

The fix is Application Audio Capture. It’s in OBS Studio 28 and later, on Windows 10 (version 2004 or later) and Windows 11:

  1. In the Sources dock, click + and choose Application Audio Capture (BETA).
  2. Name it after the app, so you know which fader is which.
  3. Pick the app’s window from the drop-down.
  4. For most apps, match by window type. For apps whose window title keeps changing, like Spotify or Chrome, match by executable name.
  5. Repeat for each app you want on its own fader.

OBS 30.1 and later have a shortcut for games. Add a Game Capture or Window Capture source and tick Capture Audio (BETA).

Then open Settings > Audio and set the global Desktop Audio to disabled. OBS’s own guide warns about this one. If you leave it on, the same sound gets captured twice and comes out as an echo.

I couldn’t confirm the current layout of TikTok LIVE Studio’s own mixer from TikTok’s help center, so the source steps here are OBS’s. Step 1 still applies to LIVE Studio, because Windows treats it like any other app.

3. Set levels by the meter colours

The meter in OBS’s mixer has three colour zones, and OBS’s docs say what goes in each:

  • Your voice: the top of yellow, just reaching the bottom of red.
  • Game sound: the lower part of yellow.
  • Music and alert sounds: green.

Red means you’re close to clipping, so nothing should stay there. The number next to each fader is that source’s level in decibels. If a quiet source is still too low with its fader all the way up, go to Edit > Advanced Audio Properties, where you can set a fader above 100%.

Viewers came for you, not the boss fight’s soundtrack. If your voice and the game sit in the same zone, the game wins.

4. Make the music duck when you talk

Ducking means one source turns down by itself while another is making sound. In OBS, you do it with the Compressor filter and a sidechain.

  1. Add a Compressor filter to the source you want turned down. That’s your music source, or Desktop Audio if you kept it.
  2. Set Sidechain/Ducking Source to your microphone.
  3. Start from OBS’s suggested ducking values: Ratio 32:1, Threshold -36 dB, Attack 100 ms, Release 600 ms, Output Gain 0 dB.

The filter’s defaults (10:1, -18 dB, no sidechain) aren’t meant for ducking, so change them. Threshold sets how far the music drops. Attack and Release set how fast it drops and how fast it comes back. With a sidechain, leave Output Gain at 0.

5. Listen to what viewers hear

Each entry in OBS’s mixer has a monitor button. It plays that source through the monitoring device you chose in Settings. You can also turn on monitoring for each source in Edit > Advanced Audio Properties.

Wear headphones for this, so the monitored sound doesn’t loop back into your mic. Talk over a song for thirty seconds and listen for the music to dip.

Want a separate mix for your headphones?

Sometimes you want Discord loud in your ears but quiet on stream. That takes two mixes, and OBS’s mixer only gives you one.

Elgato’s Wave Link 3.0 handles this. It runs on Windows and macOS (on PC it needs Windows 11), and it works with any mic or interface your computer recognises, so you don’t need a Wave mic. You can add Personal, Chat and Stream mixes out of the box, up to five in total, and each one has its own level for every source. Its driver makes those mixes show up as inputs in other apps, so you pick the stream mix as the audio input in your streaming app. I couldn’t confirm the exact device name Wave Link 3.0 shows there.

Still not right?

  • The meters don’t move at all: OBS’s quick start guide says to open Settings > Audio and pick your devices yourself.
  • One app is silent: go back to step 1. It’s usually a muted app, a slider that’s down or the wrong output device.
  • Your voice is too quiet even at full fader: add a Gain filter to the mic. It boosts quiet sources.
  • Boosting brings up hiss: add Noise Suppression. RNNoise is the default and sounds better, but it uses more CPU. NVIDIA Noise Removal only appears once NVIDIA’s Broadcast SDK is installed.
  • An echo started after step 2: Desktop Audio is still on alongside your app captures.

If you’d rather have your levels in the same app as your TikTok alerts, Campfire Studio has a sound and mixer panel that shows levels for every source.

Ducking fixes the volume, not the rights. The music you play on a LIVE still has to be music you’re allowed to use.