saycutEditing notesTry the sample editor

Published

How to Balance Dialogue and Background Music in a Video

If music makes a speaker difficult to understand, do not try to make both tracks equally prominent. Find the passages where the words matter, lower the music there, and listen where speech starts and stops. No single level setting fits every voice and song. Use playback to judge the balance rather than treating a number as the finished result.

Decide when the words should lead

Editorial recommendation: identify the words a viewer needs to catch in an explanation or interview. Make speech the focus in those passages. Music can take a more noticeable role in an opening or a shot without speech if that suits the video; the balance need not stay fixed from beginning to end.

Soundstripe describes keeping background music below speech as a starting point for dialogue-led video, but cautions that there is no universal recipe for track levels. A speaker may get quieter between sentences, or an instrument may compete with one particular word. A track setting alone cannot tell you whether the message comes through.

Listen to the source voice before adding music

Play the voice on its own first. Mark quiet sentences, sudden laughter, missing words, and passages that already sound distorted. If the recording clipped when a loud sound exceeded its available headroom, lowering that passage in the edit will not restore the clean recording. Check for another usable recording before trying to work around the problem.

If your editing setup has meters, use them to check loud peaks, but also listen for intelligibility. Descript discusses checking levels during recording, mixing, and export, and recommends listening on more than one playback device. Speech that seems balanced on headphones may come across differently through laptop speakers.

Lower music during speech, then preview the boundaries

Use this sequence: confirm that the speech makes sense without music; add the music; lower it during important spoken passages; replay each point where the speaker begins or finishes. Editorial recommendation: if one sentence is masked, locate that sentence instead of repeatedly changing the level of the entire video. Check that the music does not return with a distracting jump.

Saycut prepares a first edit from footage you upload. You can request music and audio-level changes in chat, then preview the result. A request such as “Keep the music quieter during the explanation” names both the passage and the goal. Do not assume it sounds exactly as intended: listen to the preview and identify any boundary that needs another adjustment.

Labeled example: music under a cooking explanation

Hypothetical example: a 30-second cooking video moves from an opening shot to a spoken list of ingredients, then to a finished-dish shot. Let the music be heard in the opening, lower it before the ingredient list begins, and reassess its presence after the speaker finishes. This illustrates an editing decision, not a measured result or a level that guarantees clear speech.

You could request: “Lower the background music while the ingredients are named, then let it be heard naturally over the finished-dish shot.” Preview the move from the opening to the first word and the move from the explanation to the final shot separately. If music masks the first syllable or surges at the end, point to that boundary when requesting a revision.

Pre-export listening checklist

Play the full version you plan to export, with music on. Can you understand every essential word? Does a quiet sentence lead into an uncomfortably loud laugh? Does music cover the start or end of speech? Do any passages sound harsh or distorted? Does the balance still work through headphones and another playback device? Note the location of each problem and fix that passage first.

Editorial recommendation: do not judge the whole mix by the few seconds you edited most recently. Music and a simple level reduction cannot be counted on to repair distortion already present in the recording. Both Descript and Soundstripe discuss peaks and the value of checking the listening experience. Treat the finished playback, not one target number, as your final test.

Common questions

How many dB below dialogue should background music be?

There is no difference that works for every recording. Lower the music in important spoken passages and check whether the words remain clear. If you use a number as a starting point, reassess it against the particular voice, music, and playback device.

Can I fix distorted dialogue by turning it down?

If the recording clipped, lowering its level will not normally restore the missing clean sound. Check for another usable recording first. If there is none, identify which words are affected before deciding how to handle the passage.

Sources

  1. Audio Levels for Video Editing: Where to Set Them
  2. How to Perfect Your Audio Levels For Video (Advice from Engineers)

Have a suggestion? Let us know!