9 February 2026 | 6 min read
Room tone: what it is, how to record it, and why editors beg for it
Thirty seconds of nothing is the most reusable thing you will record all day. Here is how to get it right.
Room tone is the sound of a location with nobody performing in it. Not silence: the hum of a fridge, distant traffic through a window, the particular reverberance of a corridor. Dialogue editors use it as glue, and when it is missing, every edit in a scene announces itself.
Why it matters more than it sounds
A dialogue scene is stitched from many takes recorded across hours. Each take has a slightly different noise floor. Cutting between them produces audible steps in background level, and the audience hears the edit even if they cannot name what they heard.
Room tone fills the gaps, smooths those steps and lets an editor lengthen a pause or clean a cough without leaving a hole. It is also the reference a noise reduction tool needs to learn what to remove.
How to record it
- Record at least 30 seconds, ideally 60, for every location and every significant setup change.
- Keep the same microphone, the same position and the same gain as the takes it will support.
- Everyone present stays still and silent, including the people who think they are being quiet.
- Do it before wrap, while the set is still dressed and the air conditioning is in the same state.
- Slate it verbally at the head, not the tail, so it is findable later.
The mistakes that make it useless
Recording room tone after the crew has started striking gives you the sound of a different room. Changing gain between takes and tone makes the level mismatch worse rather than better. Recording it on a different microphone gives you a different tonal signature that will not blend.
The other classic failure is recording tone on only one channel of a multitrack setup. If the scene was covered with boom and two lavs, the editor needs tone for the same combination, because each source has its own noise character.
Delivering it to post
Keep room tone in the same format as the rushes: same sample rate, same bit depth, same channel layout. If your recorder writes poly files, the tone will be poly too, and post will need it split into the same named mono stems as the dialogue.
Split it with the same tool and the same naming convention you use for the takes. Boom tone next to boom dialogue, Lav1 tone next to Lav1 dialogue. That consistency is what lets an editor drop it in without thinking.
A quick quality check
Before you move on, listen back. If you hear a phone vibrate, a chair creak or someone whisper, record it again. A single transient makes a whole tone pass unusable for looping, which is exactly how editors use it.