Music ducking lowers a background track while narration speaks, then allows it to rise when space returns. The result should support the story without making the music audibly jump after every word. Establish a clear voice and a suitable music arrangement before choosing manual or automatic ducking.
Balance the voice before involving the music
Listen to the narration alone and correct distracting level changes between clips. Use clip gain for local differences and consider compression only when the voice's dynamics need broader control.
Then add the music at a restrained level. Check whether the arrangement competes with important consonants or phrase endings. A busy melody or vocal lyric can draw attention even when its meter reading looks modest.
If a different section of the same music works better, use that editorial choice before building elaborate automation. Ducking is one part of the mix, not a guarantee that any soundtrack will fit any narration.
Decide where the music should tell the story
Mark places where music can rise meaningfully: an opening, a visual montage, a transition, or an ending. Keep it more stable beneath connected speech. A breath between clauses is usually not a new musical moment.
Before: the music surges after “Choose the folder,” then drops again for “and open the first file.”
After: the music stays beneath the complete instruction and rises only after the action is established.
The second version follows the thought rather than the waveform's smallest gaps. This is the central editorial decision whether you automate by hand or use an effect.
Use automatic ducking as a starting point
Audacity's Auto Duck documentation describes reducing selected tracks based on a control track. Its timing options determine how the music falls, rises, and responds to gaps.
Follow the instructions for the version and editor you use. Route or select the music as the material to reduce and the narration as the relevant control. Verify the result with a short passage before applying the effect across a long project.
Choose settings by listening to the actual phrase structure. If the music rises inside every sentence, change the pause behavior or use manual edits. If it falls too late, the first word can become obscured even though the rest of the sentence is clear.
Refine fades around entrances and exits
Let the music move down smoothly around the start of speech. Keep the first consonant intelligible. At the end, listen for whether the music returns with purpose or seems to leap into the space.
A fade that works for a sustained background may feel awkward on a strong rhythmic accent. Move the musical edit or adjust the envelope in context instead of treating the narration track as the only timing reference.
Use an external editor for these operations. Keep automation editable when possible, since a later script change can move the point where the music should rise.
Test on ordinary playback devices
Listen on headphones, then on a simple speaker at a modest volume. The key words should remain understandable without concentration. Check mono playback if your audience may hear the work that way.
Review the opening and ending separately. A music introduction that feels comfortable in isolation can become too loud compared with the first spoken line. An ending can also expose a sudden change that was hidden during speech.
For VocalCopyCat projects, create and approve the narration in the voice studio, download it, and build the music mix externally. Use music you are authorized to include in the intended release.
Keep a voice-only export and an editable mix. That lets you revise the script, adapt the soundtrack, or supply an alternative delivery without trying to separate the finished audio later. Approve the mix by whether the voice remains clear and the music's movement feels intentional.
