Podcast and Spoken-Word Audio
How to add intro music and duck it under your voice in REAPER
Published August 17, 2026
Put your voice and your music on separate tracks, then lower the music whenever you talk. REAPER gives you two stock ways to do this. The simplest is a volume automation dip you draw on the music track, which gives you exact control over the intro and outro. The hands-off way is automatic ducking with ReaComp, REAPER’s stock compressor, using a sidechain so the music drops by itself when you speak. Aim for the music sitting a few dB under your voice while you talk and coming back up in the gaps. Both methods use only what ships with REAPER.
An intro-music bed sounds great on its own, and then you start talking and the two fight each other. The music covers your words, or you end up dragging the music fader up and down by hand for the whole episode. The real fix is ducking: the music automatically or deliberately drops under your voice and lifts back up in the pauses. It is what makes a home podcast intro sound produced instead of homemade.
This guide covers both stock ways to duck in REAPER, when to use each, and what ducking will not do for you. You do not need to buy a plugin.
Set up two tracks first
Ducking only works when the voice and the music are on their own tracks. Put your spoken audio on one track and your intro or background music on another. If your music came in as one long file, trim it and add short fade-ins and fade-outs at the edges so it starts and stops cleanly.
Choose an instrumental piece with no lead vocal and not too much busy midrange. A music bed that is crowded in the same range as speech, roughly 200 Hz–4 kHz, will still feel like it is competing even after you duck it. The easiest ducking job starts with the right track.
Method 1: Volume automation (simplest, most control)
For a normal podcast intro, where the music plays alone, then you come in over a quieter bed, then the music fades out, volume automation is the clearest approach. You draw exactly what you want and it never surprises you.
Show the music track’s volume envelope
On the music track, open the track envelopes and enable the Volume envelope. You will see a horizontal line across the track that you can add points to.
Draw the dip
Add points so the music plays at full level during the intro, then drops down where your voice starts, stays down while you talk, and rises back up in the gaps or at the outro. Click to add points and drag them to set the level.
Set the depth by ear
A useful starting point is the music sitting about 6–12 dB below where it played on its own. Listen with the voice in and adjust until the words are easy and the music is still present underneath.
Automation is predictable and precise. You control the exact moment the music dips and the exact shape of the intro fade, and nothing reacts to a stray breath. For a fixed intro and outro it is usually faster than setting up a sidechain.
Method 2: Automatic ducking with ReaComp
When you talk over a bed for long stretches and hand-drawing every dip would be tedious, let REAPER duck the music for you. This uses a sidechain: the compressor on the music track listens to your voice and turns the music down whenever you speak. ReaComp, REAPER’s stock compressor, does this with no extra plugins.
Make the music track 4 channels
Open the routing for the music track and change its track channels from 2 to 4. Channels 1 and 2 stay your normal left and right; channels 3 and 4 become auxiliary channels you will not hear.
Send the voice to channels 3/4 of the music track
On the voice track, add a send to the music track and set the send’s destination to channels 3/4. Your voice still plays normally to the master; this send is just a silent copy the compressor can listen to.
Add ReaComp and set the detector input
Add ReaComp to the music track. In its settings, set the Detector input to the auxiliary input, channels 3/4. Now the compressor reacts to your voice instead of to the music.
Dial in the ducking
Start with a ratio around 4:1, then lower the threshold until the music ducks a few dB when you speak. Adjust attack and release to taste, then check it in context.
Good starting ranges, not fixed rules: ratio 4:1 or higher, threshold set so the music drops roughly 6–12 dB under speech, attack around 5–20 ms so the dip is not choppy, and release around 200–400 ms so the music eases back up rather than jumping. Slower release sounds smoother for talking; faster release brings the music back sooner. Tune it by ear with the voice playing.
The send to the music track goes to the auxiliary channels 3/4, which are not routed to the master, so you should not hear a doubled or echoey voice. If you do, the send destination is wrong: check that it is set to channels 3/4 and not 1/2.
| Situation | Best method | Why |
|---|---|---|
| Fixed intro, then you talk over a quiet bed | Volume automation | Exact control of the intro fade and dip timing |
| Talking continuously over a bed | ReaComp sidechain | Hands-off; the music ducks itself on every phrase |
| Short stingers or stabs between segments | Volume automation | You place each level change exactly where you want it |
What ducking will not fix
Ducking balances two tracks against each other. It does not solve everything around them.
- It will not rescue a busy or vocal-heavy music bed. If the track has its own singing or dense midrange, it will still feel like it is competing. Pick an instrumental bed instead.
- It will not set your final loudness for the platform. Balancing music under voice is separate from hitting a delivery target. Set that at the end. See how to reach the podcast loudness target of -16 LUFS in REAPER.
- It will not clean up a noisy or muddy voice. Hum, hiss, and room noise are their own fixes. See how to reduce background noise in a podcast recording.
- Automatic ducking reacts to every pause and breath. If that pumping is distracting, lengthen the release or switch that section to volume automation.
Done well, ducking makes your intro feel produced and keeps every word clear over a bed. It cannot make a mismatched song sit politely under speech, and it is not a substitute for choosing the right music and setting your final levels. Used with those, it is the piece that ties a home podcast intro together.
If it still sounds off
Quick troubleshooting order when the ducking is not behaving:
- The music never ducks. The detector input is probably still on the music itself. Set ReaComp’s detector input to the auxiliary channels
3/4, and confirm the voice send’s destination is channels3/4. - It pumps or breathes. The dip is too aggressive or too fast. Raise the threshold, lower the ratio, or lengthen the release so the music returns smoothly.
- The voice sounds doubled. The sidechain send is leaking to the master. It must feed only channels
3/4, which are not heard. - The dip is too slow to catch the first word. Shorten the attack so the music drops as soon as you start speaking.
Keep voice and music on separate tracks and lower the music under speech. Draw a volume automation dip for a fixed intro and outro, or set up ReaComp sidechain ducking when you talk over a bed for long stretches. Start with the music about 6–12 dB under your voice, judge it with the voice playing, and choose an instrumental bed so the two are not fighting in the first place.
Sources
- Reaper Accessibility Wiki, “Sidechaining with ReaComp” - step-by-step routing (4-channel track, send to channels 3/4) and setting ReaComp’s detector input to duck podcast music under voice. reaperaccessibility.com
- Cockos, “ReaPlugs” - ReaComp described as the compressor bundled with REAPER. reaper.fm/reaplugs
- Cockos, REAPER User Guide - track channels, sends and routing, track volume envelopes, and ReaComp’s detector (sidechain) input. reaper.fm/userguide