Room Tone: The Cheapest Fix in Post
Answers what is room tone and do I need it
Take a finished film that sounds amateur and listen to the gaps. Not the music, not the dialogue. The spaces between the words.
In a film without room tone those spaces are silent, and they change character every time the cut moves to a different camera or a different angle. The ear reads that as a seam. Nobody can say what is wrong, only that it sounds edited.
The fix costs thirty seconds on the day and nothing at all in post.
What to record
At the end of a take, before anyone moves, hold the room for thirty seconds. No talking, no chairs, no pages turning. Just the space.
Do it three times: once in each location, and once with the air conditioning wherever it normally sits. That is it.
If the shoot has already happened and there is no room tone, it is still recoverable. Any clean thirty seconds from the same location will do, even if it is between takes, even if someone coughs in the middle, as long as there is enough unbroken noise around it.
What it is for
Filling the gaps. Lay a bed of it across the whole scene at a low level, so there is a continuous floor under everything.
Covering the cuts. When you cut between two microphones or two angles, their noise floors do not match. The bed hides the join.
Placing a line. A line of narration or ADR with no room around it sounds like it was recorded in a cupboard, because it was. A little bed puts it back in the scene.
Smoothing an edit inside a sentence. Cutting a word out of a take leaves a click of different noise. Fill it.
How loud
Quieter than feels right while you work on it, and audible in the finished mix if you listen for it. If you can hear it as a thing, it is too loud. If you can hear the silence when it stops, it was doing its job.
The practical test: mute the bed and play ten seconds of a scene. If any cuts become audible, it was working.
The four other things that make dialogue sound recorded
Room tone is the cheapest but not the only one.
A high-pass. Eighty to a hundred hertz on every voice track, to stop the rumble fighting the music.
Consistent levels. Not levelled to one number, ridden through each line, because the last words of a sentence go quiet and that is where the meaning usually is.
A little smoothing. A gentle compressor, two or three decibels, not the aggressive setting that flattens consonants.
The right music level. Under the speech, ducking when the speech starts. Louder than you think after the fourth listen, because you know what is being said.
If the sound is the part of your film you keep coming back to, it is worth an hour of someone else's ears before more money goes into picture.