7 min read
Audio mixing for picture: balancing dialogue, effects and music
Mixing for picture is not music mixing. Dialogue leads, everything else negotiates for the space left over.
A film mix has one non negotiable rule: the audience must understand the words. Every other decision, the size of an explosion, the warmth of a score, is made around that constraint.
Work in this order
- Clean and level the dialogue first, scene by scene, before adding anything else.
- Add ambience so the scene has a floor, then place hard effects against picture.
- Bring foley in to give bodies weight and movement.
- Music last, riding under dialogue rather than fighting it.
- Check the whole reel at one fixed monitoring level, never on the fly.
Monitoring level is the real instrument
If you change the monitor volume during a mix, your judgement moves with it. Set a reference level, note it, and keep it. Quiet monitoring makes you push dialogue too loud; loud monitoring makes you bury it.
Frequency space, not just volume
When music and dialogue collide, cutting a few decibels around 2 to 4 kHz in the music usually opens a hole for speech without anybody noticing the music got smaller. Reaching for the fader is the blunt version of the same idea.
Loudness targets you must respect
Broadcast and streaming deliveries specify integrated loudness, typically minus 23 LUFS in Europe and minus 24 LKFS in North America, with a true peak ceiling around minus 2 dBTP. Cinema mixes work to a monitoring calibration instead. Measure over the whole programme, not per scene, and never normalise a finished mix by peak alone.
Deliver stems, not just a master
Every distributor eventually asks for separated dialogue, music and effects, plus an M and E for foreign dubbing. Print them from the same session at the project sample rate and bit depth, check each one for clipping and confirm the channel layout matches the delivery spec before you send anything.