Choosing the right microphone for the situation is as important as choosing the right lens. Each type has a pickup pattern that determines where it captures sound and where it rejects it.

Audio levels in production are measured in dBFS (decibels relative to full scale). 0 dBFS is the absolute maximum — any signal above it clips and distorts irreparably. Unlike visual overexposure, which can sometimes be partially recovered, clipped audio is destroyed.

The standard target for dialogue recording is -12 to -18 dBFS on your meters. This gives you headroom for unexpected peaks (a shout, a laugh, a door slam) without clipping. Peaks should not exceed -6 dBFS. It is always better to record slightly quiet (you can boost in post with minimal noise penalty) than slightly hot (clipping cannot be undone).

Room tone is the ambient sound of a location with no one speaking or moving. Every room has a unique sonic fingerprint: the hum of HVAC, the buzz of fluorescent lights, distant traffic, wind. Recording 30–60 seconds of room tone at every location is essential. Sound editors use room tone to fill gaps between dialogue edits, smooth transitions, and create a seamless audio bed. Without it, cuts between dialogue lines have jarring jumps between silence and the room’s natural ambience.

Record room tone at the end of each setup or location. Have everyone on set freeze in place and stay silent while you roll. It takes one minute and saves hours in post.