Ducker
Automatically lower one audio source when another is active - music that steps back when you speak.
intermediatePro
A Ducker lowers the volume of one signal whenever another signal is loud enough. Music drops when you talk, then comes back when you stop.
A Ducker node with its settings panel open.
ducker-expandedWhen to use it#
- Music ducks under your voice while streaming
- Game audio drops when a Discord call comes in
- Backing tracks step back for announcements
How it works#
The Ducker has three ports:
| Port | Carries |
|---|---|
| In | The audio you want to duck - music, game audio |
| Key | The trigger that causes ducking - your microphone |
| Output | The ducked result |
The Key signal is only listened to. It is never mixed into the output - it just decides when to pull the In signal down.
The mic still needs its own separate connection to the stream. The Ducker only outputs the ducked music.
Setting it up#
- Right-click the canvas → Mixing & Effects → Ducker.
- Connect the audio to duck (music) to In.
- Connect the trigger (mic) to Key.
- Connect Output onward to your destination.
- Connect the mic separately to that same destination, so it is actually heard.
Parameters#
Open the node's Settings to reveal them.
| Parameter | Range | Default | What it does |
|---|---|---|---|
| Threshold | -60 to 0 dB | -20 dB | How loud the Key must get before ducking starts |
| Duck | -60 to 0 dB | -20 dB | How far the In signal drops. -60 is near silence |
| Attack | 1-200 ms | 10 ms | How fast the duck engages once triggered |
| Release | 50-2000 ms | 250 ms | How fast level returns after the Key goes quiet |
| Hold | 0-1000 ms | 100 ms | How long to stay ducked after the Key drops below threshold |
| Fade | 10-2000 ms | 400 ms | Smoothing on the level change |
The node header summarises the two you change most: duck -12 dB · thr -30 dB.
Tuning it#
| Symptom | Fix |
|---|---|
| Quiet speech never ducks | Lower Threshold |
| Keyboard clicks duck the music | Raise Threshold |
| First syllable gets lost | Lower Attack to 5-20 ms |
| Music pumps between words | Raise Hold to 200-400 ms |
| Music returns abruptly | Raise Release; 300-800 ms feels natural |
Set Threshold first so normal speech triggers reliably and room noise does not, then Hold.
A starting point for voice over music#
| Parameter | Value |
|---|---|
| Threshold | -30 dB |
| Duck | -15 dB |
| Attack | 10 ms |
| Release | 400 ms |
| Hold | 250 ms |
| Fade | 100 ms |
Limits and gotchas#
- Pro only.
- The Key signal is never passed through - route your mic separately as well.
- Ducking is one-directional. For mutual ducking you need two Duckers.
- Very low Duck values (
-50 dBand beyond) sound like a hard mute; most broadcast ducking sits between-9and-18 dB.
Related#
- Volume Knob - manual level control
- Streaming Setup - a full setup using a Ducker
- FX Module - if you want a plugin compressor with sidechain instead