Background music under a voiceover

Background music under a voiceover is the wrong job for this catalog. The tracks are vocal rap. Your voice plus that voice is two people talking. Ducking the song until the lyric disappears leaves a muffled chorus and a worse video.

What to do instead

Talk in silence, or talk over a real instrumental bed from a source licensed for that. Use an Arctaholic track only where the voiceover stops: the open, a b-roll cut, the end. Vlogs get the longer version of this cut in copyright free music for vlogs. Commentary gets it in commentary videos.

The license would still allow you to bury the track. The license is not a mix note. “Allowed” and “sounds bad” are different.

Do not “fix” the collision by stripping the vocal out. That is a remix, and it is not allowed. See can I remix and vocal or instrumental.

If you can hear both voices fighting, the mix is telling you to cut one of them.

Background music under a voiceover, and the cut that saves the take

Background music under a voiceover is the placement this catalog loses. Two voices at once make the viewer work. They will not work. The fix is structural, not a fader trick.

Record the voiceover first, with no music. Then drop the track only into the gaps you already have: before the first word, between two thoughts, under a b-roll cut where you stopped talking, and after the last word. If the script has no gaps, do not invent a bed. Rewrite the open so the first two seconds are picture, then start talking in silence.

Sidechain ducking and “vocal remover” tools are the wrong repair. Ducking a rap vocal leaves consonants pumping under your voice. Removing the vocal is a remix. The license allows a trim for length. It does not allow a new instrumental made from the master.

Podcasts and commentary videos are the same problem with a different runtime. The podcast guide and the commentary guide use the same gap rule. A claim does not appear because you ducked the song. It appears because a system matched the audio. YouTube’s claim help describes the match. The license describes the permission.

Record the voice with the song off, always. If the song was playing in the room, it is in the take and you cannot trim it out without damaging the words. Re-record the sentence. Then build gaps on purpose: two seconds of the place before you speak, a cutaway with no voice, two seconds after the last line. Drop the track only there. Export and listen on earbuds, because earbuds make a buried vocal more obvious than desk speakers do. If you still hear two people, the gap is too small. Widen it or delete the track. Do not reach for a vocal-remover tool. If two voices still collide, delete the track from that stretch. Record the voice with the song off, then add the track only in the gaps.

Questions

What if my voiceover is only ten seconds?

Then a sting before or after those ten seconds is enough. You do not need the song under the sentence.

Are podcasts the same problem?

Yes, under the interview. Intros and outros are the fit. See podcasts.