How do I make a voiceover sound as if it was recorded in the scene?

Wire the voice through Audio FX with the Room (indoor) effect and a Wet / Dry mix around 25 to 35. Then merge the result onto the video.

Which effects does Audio FX have?

Nine reverb spaces, from Room and Bathroom to Church and Arena, plus Telephone, Megaphone / PA, Echo / Slap-back and a Custom mode with delay and EQ.

How many credits does Audio FX cost?

A flat 20 credits per run. If a reverb cannot be calibrated safely, the run fails and the credits are refunded.

Should I add the effect before or after merging the voice onto the video?

Before. Put Audio FX in front of Merge Video & Audio, so that the placed voice is what lands on the video.

Audio FX node — reverb, telephone, megaphone and echo effects for voices

Put a voice in a room, a car or a church with a reverb, or add a telephone, megaphone or echo effect. A flat 20 credits per run, plus custom delay and EQ.

The Audio FX node puts an effect on an audio clip. It adds a reverb that places a voice in a space, a telephone or megaphone sound, an echo, or a custom delay and EQ. Its main use is to place a dry voice in a scene. A clean Text to Speech voice over a video can sound as if it was recorded somewhere else. A room reverb makes it sound recorded in that space.

When to use it

Make a generated voiceover sound as if it was recorded in the room on screen.
Make a line sound like a phone call or a public address announcement.
Add an echo or a slap-back to a voice.
Add an effect to a single stem from Audio Separation.

Quick start

Add the node

Press Tab on the canvas and choose Audio › Edit Audio › Audio FX.

Wire the audio

Wire an audio node into the Audio input: a generated voice, an upload or a separated stem.

Choose the effect

Open the settings panel and choose an Effect. The default is Room (indoor). For a reverb, set the Wet / Dry mix.

Run it

Click Run on the node. The processed audio appears on the node with a player.

A dry generated voice gets a room reverb, so it sounds recorded in the scene it is merged onto.

Input and output

Input	Accepts	What it does
Audio	Audio nodes, such as Text to Speech, Upload Audio and Audio Separation	The audio to process.

The output, Audio, is the processed audio. It connects to any audio input, such as Merge Video & Audio, Voice Changer or Mix Audio.

Settings

Setting	What it does
Effect	The effect to apply. The default is Room (indoor).
Wet / Dry mix	Reverb effects only. From 0 to 100. Higher means more room and less direct voice. Each reverb has its own default.
Delay (ms)	Echo and Custom only. From 20 to 2,000. The default is 250.
Decay	Echo and Custom only. From 0.1 to 0.9. The default is 0.4. Higher gives more repeats.
EQ Low (dB) and EQ High (dB)	Custom only. From -20 to 20. The default is 0.

Every effect

Reverb spaces place the voice in a space:

Effect	Space
Room (indoor)	An ordinary room. For dialogue.
Bathroom	Small, bright and tiled.
Car interior	A tight, damped cabin.
Hall / Lobby	A medium-sized room.
Concert Hall	A large space with a lush tail.
Church	A cathedral: long and dark.
Cave	Dark and diffuse.
Arena / Stadium	A huge space with a bright public address tail.
Outdoor (open air)	Almost dry, with faint reflections.

Character effects change the sound itself:

Effect	Sound
Telephone	A band-limited phone line.
Megaphone / PA	A mid-forward sound with some grit.
Echo / Slap-back	A distinct repeat of the sound.
Custom	Your own delay and EQ.

Credits

Audio FX costs a flat 20 credits per run.

Reverb effects calibrate their level when the node runs. If the calibration cannot complete safely, the run fails and the credits are refunded: the node never renders a reverb at a guessed level. A passing failure succeeds on a retry. If the same run keeps failing with a calibration error, report it instead of retrying.

Tips

Keep dialogue reverb light. For dialogue indoors, Room (indoor) with a mix of about 25 to 35 is usually enough. Too much reverb makes speech muddy.
Match the reverb to the shot. Use Church or Cave for big stone spaces, Car interior for car interiors and Outdoor (open air) for exteriors, kept subtle.
Apply effects before the merge. Put Audio FX in front of Merge Video & Audio, so that the placed voice is what lands on the video.
Several voices at once. Voice Changer Pro has the same effects built in, applied to the new voices before the background music is mixed back.

From the API

POST /v1/audio-fx takes audioUrl and the optional preset, mix, delayMs, decay, eqLow and eqHigh. The MCP tool is apply_audio_fx. See Voice and media.

Audio FX