To mix voiceover and background music, start with a clear spoken track, add an instrumental bed at a supporting level, and lower the music when the narrator speaks. Then check the beginning, ending and actual duration before placing the soundtrack under your video.
BrandGene can generate both tracks in one audio package and mix completed tracks in the browser. It also keeps the originals available for download. This guide covers that workflow and the decisions that matter when the first mix needs a change.
Listen to the originals before changing the mix
Play the voice on its own. If the brand name is wrong or the copy runs too long, a quieter music bed will not fix it. Revise the voiceover or script first.
Then play the music alone. Listen for vocals, a dominant melody, a dramatic change in the middle or an abrupt ending. If any of those conflict with the ad, request a simpler instrumental arrangement before trying to balance it against speech. The background music guide includes directions for narration-friendly tracks.
What ducking does
Ducking lowers the backing music while the voice is active. The goal is to keep the words easy to understand while retaining a sense of musical continuity. It is useful for short commercial reads because the offer and action need to stay intelligible.
A quieter music setting and ducking solve related problems. The backing setting controls the music's general presence; ducking creates additional room under narration. Neither can fully rescue an arrangement whose lead instrument constantly competes with the voice.
For a first pass, choose softer backing if the ad contains detailed product information. A more present bed can suit a sparse script with long visual moments between phrases. Judge the result on the actual track, rather than assuming a named level will fit every voice and genre.
Create a mixed soundtrack in BrandGene
1. Request a complete package
Inside Image Agent's audio workspace, describe the product, desired voice and instrumental backing. The agent plans the required tracks and shows the total credit quote. Confirm the package after checking the transcript and music direction.
Create a voiceover and instrumental music package for [product].
Transcript: "[approved ad script]"
Voice: clear, warm and conversational in [language].
Music: restrained [mood and instruments], with no vocals and room for speech.
Target use: a 15-second product video; prioritize finishing the spoken action clearly.
Show the track settings and total credit quote before generating.
For help writing that brief, start with the AI audio ad generator guide.
2. Let both tracks finish
The workspace needs the completed original tracks before it can mix them. Keep it open for browser processing. If you close the page before mixing finishes, return to the session to let the workspace finish the missing mix.
An unsupported audio decode or browser access problem can prevent mixing. Use the available retry when appropriate, or download the originals and combine them in an external audio or video editor. A failed mix does not mean you need to generate the voice and music again.
3. Choose a backing level
Use soft, balanced or present backing to change how prominent the music feels. For the first review, focus on whether each spoken sentence remains clear through a phone speaker. Listen again at the final action, where music or an ending flourish can mask the last words.
4. Download and align with the footage
The completed browser mix is a stereo WAV. Download it and import it into your video editor. Align the start of the narration with the intended scene, then check the ending against the picture cut. The audio workflow does not automatically merge it into a video file.
Revise one part without replacing the other
| What needs changing | What to ask for | Generation cost behavior |
|---|---|---|
| Voice pronunciation or script | Keep the completed music; generate a revised voice | The new voice is quoted |
| Music mood or arrangement | Keep the completed voice; generate new instrumental backing | The new music is quoted |
| Music is too prominent | Reuse both completed tracks with softer backing | Mixing adds no credits |
| Mix failed but originals play | Retry mixing or finish externally | No new track generation is required |
Reuse applies to completed tracks in the same session. Ask the agent to keep the approved component rather than starting an unrelated new session. Review the new package quote to confirm that only the intended track will be generated.
A useful revision request is: “Reuse the completed voice and music from this session. Make the backing soft and create a new mix without generating either track.” If you want different music, state that separately so the plan includes the new generation.
Check duration after mixing
A 15-second backing track does not force the voice to end at 15 seconds. BrandGene's mix preserves narration rather than silently cutting it off. Speech that overruns, plus any intro or tail, can make the mix longer than the target ad.
If the destination has a strict time limit, shorten the script or adjust the edit before exporting the final video. Listen through the full spoken action after every timing change. Avoid solving a timing problem by cutting the last word.
For longer material, the browser mix currently has a 180-second and 50 MB limit. Keep longer production work in an external editor and use the original tracks as your source material.
A final listening pass
Check on a phone speaker and headphones. The phone reveals whether the message survives a small speaker; headphones make abrupt cuts and unwanted artifacts easier to hear. If the mix only works at high volume, try lowering the backing before increasing the overall output.
When evaluating an ad version, keep the offer and visuals fixed while changing one audio variable. Save the script, voice direction and music brief with the version so you can reproduce the creative decision later.
Frequently asked questions
Does mixing use more credits?
Browser mixing and reuse of completed same-session tracks add no generation credits. Creating a new voice or music track is a separate quoted generation.
Can I upload any two recordings into this workflow?
This workflow mixes supported completed tracks from your BrandGene session. For arbitrary external recordings, download your generated track and finish the combination in an external editor.
Can I download the separate voice and music?
Yes. The originals remain available alongside a successful mix, including when you want to do the final timing and level changes elsewhere.
Does the mix include captions or video export?
No. It is an audio download. Add it to your footage and create captions in your video editor as needed.