ADR

Automatic Dialogue Replacement (ADR) — also called looping — is the process of re-recording dialogue in a studio after production, synchronized to the picture. The term 'automatic' is historical: early systems used a film loop that played the same few seconds repeatedly, giving the actor multiple attempts. Modern ADR sessions use Pro Tools with a video track synced to the audio timeline. The actor watches the performance on screen and re-speaks the lines, matching lip movements and emotional delivery. The ADR workflow starts with a spotting session where the supervising sound editor, dialogue editor, and director review every line of dialogue and flag takes that need replacement. Flags fall into categories: technical (generator noise, wind, clothing rustle on the lav), performance (the director wants a different read), and production (the line changed after the shoot). A typical feature generates 200-600 ADR cues. Each cue is logged with timecode, the original line, the replacement line (if changed), and a priority flag (must-have vs. nice-to-have). Mic matching is critical. If the production audio was captured on a Sennheiser MKH 50 shotgun, the ADR should be recorded on the same mic — or as close as possible. If the production used a Sanken COS-11 lavalier, record ADR with a lavalier in the same position. The reason: each microphone has a unique proximity effect, off-axis coloration, and self-noise signature. When the mixer crossfades from production audio to ADR within a single line, a mic mismatch creates an audible timbre shift that pulls the audience out of the scene. Recording ADR in a treated vocal booth (RT60 under 0.3 seconds, ambient noise below NC-20 — the long-standing noise-criterion standard for professional recording and broadcast studios) ensures clean, dry recordings that the mixer can later match to the production room tone.

Not every problematic line needs ADR. iZotope RX 11, the current industry-standard dialogue-repair suite, can rescue a surprising amount of production audio. RX Spectral Repair can remove steady-state tones (HVAC hum at 60Hz, generator whine at 120Hz) without affecting dialogue. RX Voice De-noise handles broadband noise (traffic, air conditioning, room rumble) down to about 12-15 dB of reduction before artifacts appear. RX De-reverb reduces room reflections, though heavy reverb in production audio is still a hard limit — no algorithm can fully remove it. The decision tree: if the dialogue is intelligible and the problem is steady-state noise, try RX first. If the dialogue is intelligible but the noise is impulsive (hammering, dogs barking, overlapping crew chatter), RX can reduce but not eliminate it — partial ADR (replacing only the affected words) may work. If the dialogue is unintelligible or the performance needs changing, ADR is the only option. ADR performance matching is the hardest part. Actors deliver differently in a quiet booth than on set. Adrenaline, scene context, physicality — all absent. Directors often attend ADR sessions to coach performances back to the original intent. Beep systems (three beeps before the line starts, with the fourth beep marking the in-point) give the actor timing. Streamers — a vertical line moving across the screen toward the in-point — provide a visual cue. Pro Tools with the ADR module (available in Ultimate) automates cue management, take logging, and comping. Budget consideration: principal cast ADR runs $1,500-3,000 per actor per session day. Loop group (background voices — crowd, radio, phone) runs $800-1,200 per group per day.