ADR and Dubbing Voice Over Sessions: The Practical Guide

The short answer: ADR (automated dialogue replacement, also called looping) is re-recording dialogue in a studio against the finished picture, because the set recording is unusable or the line has changed. Dubbing is the same discipline in another language, with a script rewritten so the new words fit the actor's mouth. In both, a cue (three beeps, a streamer wiping across the screen, or a scrolling rythmo band) tells you exactly when to speak, and the job is to hit that window with a performance that still sounds alive.

Most voice over work starts from a blank script and a director's sense of tone. ADR and dubbing start from something that already exists: a finished performance, on screen, with its own rhythm, breath and emotional arc, and in dubbing a mouth moving in a language you are not speaking. This guide covers why the work exists, the three cueing systems you will meet, loop groups and walla, how lip sync is really judged, what the union contracts say, how a session runs, how to set up for remote ADR at home, and how actors and podcast-trained voices build the skill and find the work.

What ADR Is and Why Films Need It

Every film and series records dialogue on set, usually on a boom mic and a radio lav on each actor. A lot of it cannot be used. Wind, traffic, a generator, an aircraft, a camera dolly creaking, a costume rustling against the lav: any of these can mask a word. A French industry guide to postsynchronisation from AcademieCine lists the triggers plainly: a gust of wind, a noise covering a word, or a real problem of intelligibility. It also states the rule post teams work by, which translates as "use up the production sound before you decide to redo it". Editors hunt for the same word in other takes and on other mics first. ADR is what is left when they cannot find it.

Bad sound is only one reason. ADR also adds lines that were never shot (a new joke, a clarifying sentence over a character's back, a name change for legal reasons), replaces a performance the director wants changed after the edit, records broadcast-safe alternatives to swearing for TV and airline versions, and captures the breaths, grunts and efforts that a fight or a chase needs. When a scene is shot in a crowd, nearly everything heard in the background is added afterwards by a loop group. The craft is recognised at the top of the industry: the Motion Picture Sound Editors' 73rd Golden Reel Awards in March 2026 gave its Feature Dialogue / ADR award to the sound team on Sinners, and its broadcast long form Dialogue / ADR award to Adolescence: Episode 2.

For a voice actor, ADR work comes in three shapes. You might re-record your own on-camera lines. You might be a sound-alike covering for an on-camera actor who is unavailable, which happens in animation, games and live action. Or you might be one of a group of loopers who fill every unscripted voice in a scene. The core skill is the same in all three: matching the timing of an existing picture exactly, syllable for syllable, while sounding like a person in that moment rather than someone reading off a screen in a quiet room.

Beeps, Streamers and the Rythmo Band

The name looping comes from the original method. A short loop of film containing the line was projected over and over while the actor rehearsed and recorded against it, sometimes with a streamer or a punch on the film as a visual cue, as Simon Fraser University's sound glossary describes. Digital systems kept the idea and added two more ways of counting you in.

Beeps

The standard on US and UK ADR stages is three beeps, one second apart, followed by a silent fourth beat on which the line begins. New York re-recording mixer Jeff Hinton's ADR primer for Frame.io lays it out as beep, beep, beep, line. The beeps are usually paired with a visual mark on screen. Your job is to treat the fourth beat as the first syllable, not as a signal to start breathing. If the character takes a breath before the line, you take it on the third beep.

Streamers and punches

A streamer is a vertical line that wipes across the picture and arrives at the edge of the frame on the exact frame where the line starts, often ending in a flash of light called a punch. On film it was physically marked onto the print. Today it is drawn by cue software over the video feed: Sounds In Sync's EdiPrompt, for example, displays a cue's text, counters, streamers and progress bars, and can run several cues inside one take. Streamers help most on long or overlapping cues, where you need to see the whole run-up, not just hear it.

The rythmo band

French-language dubbing uses a different system. The bande rythmo is a strip under the picture where the words of the line scroll past a fixed bar, stretched or squeezed so each syllable crosses the bar at the moment it should be spoken. AcademieCine describes it as a kind of karaoke of the lines to be redone, with breathing marks included, and notes that the adaptor places the lip consonants P, B and M and the ends of sentences on the openings and closings of the lips. It calls the rythmo band a francophone specialty, while English-speaking studios record on loops. Software such as Mosaic, Cappella and VoiceQ (which has a built-in digital rythmo band) generates the band from the script. If you work in Quebec or France, you will read from it; elsewhere you will mostly hear beeps.

Loop Groups and Walla

On set, only the principal actors are miked. The extras in a restaurant, a precinct or a battlefield mime their talk so they do not bury the dialogue. Everything they appear to say is added afterwards by a loop group: a team of voice actors who watch the scene and improvise the background. NPR's Susan Stamberg sat in on one in 2015, recording for the series Bones on a Fox sound stage: seven loopers stood in a circle under two overhead microphones, watched a silent restaurant scene on a large screen, and on cue began to move and talk, often off-mic. She reported that 37 seconds of TV screen time could carry six layers of looping, with feature films carrying many more.

Walla is the murmur of a crowd, and the name comes from the old habit of extras muttering "walla walla". Modern loop groups do not say walla. They research. The looping coordinator in that NPR piece, Barbara Harris of The Looping Group, described having actors study street maps and local landmarks so that background talk set in Atlanta sounded like Atlanta, and having the Bones loopers learn how forensic experts actually speak. A line that pops out of the background has to be one the scene could plausibly contain, and in the right period, region and register.

Loop group work has its own technical rules too. Netflix's dubbing delivery guidelines ask studios to record walla as close as possible to the original version, to skip reactions that already exist in the music and effects track, and to bring group walla down by 10 dB in the premix so the principal dialogue sits clearly on top. That last rule tells you something about the job: walla is meant to be felt rather than heard, and a looper who keeps getting noticed is doing it wrong.

• Specific beats general. "Two coffees, one oat" is better walla than "rhubarb rhubarb". The mixer can lower a specific line; they cannot make a vague one sound real.
• Keep names and brands generic unless the coordinator says one is cleared. A real company or person audible in the background can become a legal problem.
• Match the room's energy, not your own. A funeral reception and a sports bar have different volumes, and the group adjusts as one.
• Know your place in the frame. Loopers playing people near camera step closer; people at the back of the room work off-mic. Perspective is part of the performance.

Lip Sync and Phonetic Matching

Dubbing asks for everything ADR does, plus a language gap. A translated line rarely has the same length, stress pattern or mouth shapes as the original, so a dialogue adaptor rewrites it to fit. Translation scholars break the fit into parts. Frederic Chaume's widely used model, summarised in Giselle Spiteri Miggiani's 2021 study of English dubs on Netflix in the Journal of Specialised Translation, separates isochrony (the line starts and stops when the mouth does, with the pauses in the right places), phonetic synchrony (the visible mouth shapes match the sounds) and kinesic synchrony (the words match the gestures and body language). Spiteri Miggiani adds a fourth, rhythmic synchrony: the speed of the mouth flaps.

Phonetic synchrony focuses on a small set of sounds, because those are the ones a viewer can actually see. Chaume names bilabial consonants (P, B and M, where the lips close), labiodentals (F and V, where the top teeth touch the lower lip) and lip-rounded vowels. In close-ups, he writes, phonetic equivalence can override even the meaning of the line. A wide shot or an actor with their back to camera frees the adaptor to write for sense; a tight close-up forces them to write for lips.

What does that ask of the actor? Three things. First, length is non-negotiable: Spiteri Miggiani found isochrony was the parameter English dubs respected most, followed by lip shapes, with rhythm the least kept. Second, you will sometimes be handed a line that is slightly unnatural because it has to be, and your job is to make it sound like speech anyway. Her study found the most common weaknesses in English dubs were flat speech melody, stressed syllables stretched to fill time, and a hollow, soundproof-studio sound in outdoor scenes. All three are performance problems as much as script problems. Third, English is often shorter than the source language, so you may need to fill time without sounding slow, which is harder than speeding up.

1. Watch the mouth, not the face. On your first pass through a loop, mark every lip closure you can see. Those are your anchors.
2. Land the first and last syllables. If the line starts and stops on the mouth, small drift in the middle is forgiven. Starting late is not.
3. Breathe where they breathe. A visible inhale with no sound under it, or a sound with no visible inhale, breaks the illusion faster than a missed consonant.
4. Match the effort, not just the words. If the actor is running, carrying something, or crying, your voice has to carry that strain even though you are standing still.
5. Ask about the standard. Animation usually allows more flexibility than live action, and some markets prefer natural adaptation to frame-perfect lip match. Ask what the project expects before you start.

SAG-AFTRA Dubbing and ADR Contract Basics

In the US, ADR and dubbing sit under different union agreements, and the split surprises people. SAG-AFTRA's own dubbing guidance says looping and ADR are not covered by the dubbing agreements but by the Codified Basic Agreement, the main film and TV contract, with one exception: a star re-recording their own performance in a foreign-produced picture alongside dubbing performers falls under the dubbing agreement. So if you loop English lines on an American feature, you are working under the same contract as the on-camera cast.

The 2026 TV/Theatrical contract was ratified on 4 June 2026 by 91.42 percent of voting members, according to LA Magazine, and runs from 1 July 2026 to 30 June 2030. The producers' published wage tables put the day performer minimum under the Codified Basic Agreement at $1,283 from 1 July 2026, rising 3 percent a year to $1,402 by July 2029. How your looping time is paid depends on how you were engaged on the picture: the 2023 deal let producers include one prepaid looping day in the pay of weekly performers on theatrical films earning $10,000 or more a week. A day player called back to loop is a different case, so check the call with the union before you assume.

Dubbing is hourly. The 2021 SAG-AFTRA Dubbing Agreement set $87 an hour with a two-hour minimum call and 5 percent yearly increases, time and a half for the ninth and tenth hours and double time after that. The Modified Dubbing Agreement, which covers foreign-language films and series dubbed into English or Spanish, published rates of $95, $99 and $103 an hour across its 2023 to 2026 wage table, again with a two-hour minimum. SAG-AFTRA's reference copy of its Netflix Dubbing Agreement lists $83 an hour with a two-hour minimum, counted in quarter hours, subject to later general increases. Pension and health contributions are paid on top. These are minimums; established dubbing actors and leads can negotiate more.

AI is now part of every contract conversation. The 2023 TV/Theatrical agreement required consent for most digital alterations of a performance but, as the Authors Guild summarised, carved out exceptions for post-production alterations, dubbing and foreign-language adjustments, including adjusting lips or voice to fit a dub. Reuters reported in July 2025 that SAG-AFTRA's contract permits AI-assisted dubbing from foreign languages into English provided the dubbing actor is paid. For a longer look at what those protections mean in practice, see our guide to the SAG-AFTRA AI agreement three years on, and for the basics of joining, SAG-AFTRA membership for voice actors.

How an ADR or Dubbing Session Runs, Hour by Hour

Sessions vary, but the running order rarely does. Here is what happens before you arrive, in the first few minutes, through the body of the booking and after you leave.

Before the day: spotting and cue sheets

The dialogue editor, the supervising sound editor and usually the director watch the cut and decide what to replace. AcademieCine describes this viewing as detective work, sequence by sequence, rewinding and relistening, with the mixer trying filters and plug-ins first. Every surviving line becomes a cue with a number, a character, a timecode and a reason. CalArts' 2 Pop production guide suggests budgeting an hour of preparation per minute of ADR to be recorded, and splitting the script into a separate list for each character. You will usually receive your lines, and sometimes the clips, ahead of time. Watch them, more than once, with the sound on.

The first twenty minutes: mics and context

The ADR mixer tries to match the set. Hinton describes asking for the production microphones and approximating their distance and angle, and typically recording two tracks: one boom and one lav. Netflix's dubbing guidelines similarly suggest a large-diaphragm condenser plus a boom mic, each on its own track. Then the whole scene is played with its original sound so you can hear where the character is, how loud the room is and what the other actors are giving. Ask questions here, not halfway through: what is being replaced and why, how close the shot is, and whether the director wants a match or a new reading.

The body of the session: cue after cue

Each cue follows the same loop. You hear the original line in your headphones, often two or three times. The beeps count you in and you speak on the fourth beat while watching the picture. The mixer plays the take back against picture so everyone can judge sync, and notes which takes the director prefers. A dubbing session works the same way, cue by cue and character by character, often one actor at a time, which is why a dub cast may never meet. AcademieCine notes that post teams try to leave the session with one or at most two choices per line, so the decision is made in the room. If you are asked for a version "for safety", give a genuinely different read, not the same one again.

Late in the booking: efforts, wild lines and pickups

Screams, fights, sobbing and heavy breathing are hard on the voice, so experienced directors put them near the end of the booking where they cannot ruin the dialogue that follows. This is also when wild lines get recorded, meaning added or off-screen dialogue with no mouth to match, and alternative lines for broadcast versions. Warm up for this part of the session specifically, and keep water and a plan for your voice; our vocal warm-up guide covers the routines that hold up across a long booking.

After you leave: editing and the mix

The ADR editor syncs the chosen takes to picture, often nudging individual words, and the re-recording mixer rebuilds reverb, perspective and background so the line sits inside the scene. Netflix's mixing style guide for dubbed content tells mixers to tighten lip sync where possible but to avoid time-stretching that causes artifacts, to match efforts and breaths to the original, and to mix dialogue as if it were recorded on set rather than laying words on top of the effects. If you hear your ADR in the finished film and do not notice it, everyone did their job.

Home ADR Setups and the Software Involved

Remote ADR went from rare to routine during the pandemic. In a 2020 Broadcast feature, the UK post house Boom described sending a remote ADR kit with studio-quality microphones to actors' homes, then running virtual sessions driven by an experienced ADR mixer while the production team directed from their own homes. A Production Expert case study from the same year shows the other common pattern: actor in an approved home studio, director and client listening over Source-Connect, with a local backup recording of the whole session in case the stream dropped. Both patterns are still in use, and many producers now expect a voice actor's home studio to handle at least pickups.

What a home ADR or dubbing setup needs is less exotic than people assume. Sound On Sound notes that a conventional capacitor or dynamic mic is enough for narration and ADR sessions. What matters more is matching the set, which is why it helps to have a shotgun mic available alongside your main voice mic. Sennheiser's MKH 416, introduced in 1975, is one of the most common boom mics on set, which makes it a common choice for matching. Your room must be dead and quiet, because the mixer will add the scene's acoustics later and cannot take your room away. The home studio setup guide covers treatment and noise floor in detail.

• Picture playback in sync. You must see the video with no lag against your headphones. Netflix's guidelines warn that frame rates and codecs can cause playback timing variations, so check sync before you judge it.
• Closed-back headphones so the guide track and beeps do not bleed into the mic.
• A DAW with timecode and markers. Pro Tools is the post-production default. Steinberg's Nuendo can organise ADR sessions with cycle markers per line, tagged by character and actor, as Sound On Sound's review showed.
• Cue and prompt tools if you do this often. Sounds In Sync's EdiCue builds cue sheets and Pro Tools or Nuendo sessions with a track per character; EdiPrompt draws streamers and text over the picture; VoiceQ covers beeps, streamers and a digital rythmo band; and Avid's Cue Pro, developed with Non-Lethal Applications, puts cueing and overlays inside Pro Tools 2025.6 and later.
• A remote connection the studio trusts. Source-Connect is the usual standard for high-quality streams; see our guide to taking direction remotely over Source-Connect and Zoom.
• A local backup recording. Record every take on your side as well as theirs. A dropped connection should never cost the take.

Deliver what the studio asks for, not what sounds best to you. For Netflix dubs, the guidelines call for files that match the original version's sample rate and bit depth (48 kHz, 24-bit in their naming convention), each character on a separate track, unprocessed and synced dialogue with a short fade at the head and tail of each clip, closer mic work for whispers, and no destructive processing or time-stretching unless essential. In practice that means no compressor, no EQ and no noise reduction baked in.

How to Train for ADR and Dubbing

The skill is timing without stiffness, and it can be trained deliberately. The SAG-AFTRA Foundation runs a hands-on ADR and looping workshop in Los Angeles, taught by casting director and NYU Tisch teacher Lynnanne Zager, limited to 16 actors per three-hour session, in which participants work to picture on group looping, individual sync, walla and efforts across animation and live-action material. That list is a good syllabus whether or not you can attend.

1. Sync drills, daily. Use the beep drill above on five lines a day. Score each take against picture for start, end and lip closures. Ten minutes is enough.
2. Shadow a dub. Watch a dubbed scene with the sound off and try to match the original actor's mouth in your own language. Then compare against the professional dub and notice what the adaptor changed.
3. Improvise in character. Loop group work is improvisation. Take an improv class, then practise giving five plausible, specific lines for a background person in any scene you watch.
4. Record efforts properly. Practise breaths, laughs, gasps and reactions that sound real at a distance from the mic, without hurting your voice.
5. Keep notes on every character. Pickups can come months after the session. A character bible with pitch, placement and attitude notes lets you match yourself.

Podcasters have an unexpected head start here. Anyone who has edited their own show already listens for breaths, mouth noise and the exact point where a word starts. Hosts who cut ad reads into episodes also know how long a sentence really takes. What they usually need to add is acting: committing to someone else's emotion on someone else's clock.

Getting Into Dubbing and ADR Work

The demand is real. Netflix told The Hollywood Reporter in 2019 that viewing of dubbed content on the service was growing by more than 120 percent a year, and the company has said its research found that viewers who say they prefer the original language are often more likely to watch the dub. Reuters, citing GWI, reported in 2025 that 43 percent of viewers in Germany, France, Italy and Britain prefer dubbed content to subtitles. Every foreign-language hit on a global streamer creates English dubbing work, and every English-language hit creates work in dozens of other languages.

There is pressure on that demand from AI. Amazon announced an AI-aided dubbing pilot for 12 licensed titles in English and Latin American Spanish in March 2025. When AI "beta" English dubs of anime including Banana Fish appeared on Prime Video in late 2025, the backlash from fans and voice actors was severe enough that the English versions were removed within days, Anime News Network reported. Viaplay pulled an AI-assisted German dub of a Polish series after viewer complaints, according to Reuters. The lesson so far is that audiences notice flat dubs, which is an argument for human performers who are very good at sync.

Most ADR and dubbing work is booked through relationships rather than open calls. Loop group coordinators, dubbing studios and dialogue directors hire people they know can take direction fast and hit sync within a few takes. Practical steps:

• Build a demo that proves sync. A short reel of you matching existing picture, with the original muted, shows the skill directly. Use material you have rights to, or clearly label it as a practice piece, and never post it publicly without permission.
• Contact dubbing studios directly. The big hubs include Los Angeles for English, Mexico City and Madrid for Spanish, Paris and Montreal for French, and Berlin and Munich for German. Most studios list a talent submission process on their own sites.
• Get into the room. Workshops run by working ADR directors and loop coordinators, like the SAG-AFTRA Foundation sessions, are where coordinators meet new voices.
• Show your range. Loop groups value people who can do ages, accents and specialist vocabulary convincingly. Our guide to character voices and accents for animation and games covers building that range.
• Make your home studio post-ready. Many pickups and dubs are now recorded remotely. A quiet, dead room, a shotgun mic for matching, and a reliable Source-Connect setup make you bookable for work that used to need a stage.

Frequently asked questions

What is the difference between ADR and dubbing?

ADR replaces dialogue in the same language, usually by the same actor, to fix sound problems or change lines. Dubbing replaces all the dialogue in a different language, with new actors and a script adapted to fit the original mouth movements. In the US they are also covered by different SAG-AFTRA agreements: ADR and looping under the Codified Basic Agreement, dubbing under the dubbing agreements.

Why does ADR use three beeps?

Three beeps one second apart give a steady count, so the performer can feel the tempo and start on the silent fourth beat exactly. The line starts on the right frame without anyone counting aloud. Many sessions pair the beeps with a streamer on screen.

How much do dubbing actors get paid?

Under SAG-AFTRA's dubbing agreements, union dubbing is paid by the hour with a two-hour minimum. The 2021 Dubbing Agreement set $87 an hour, the Modified Dubbing Agreement's table reached $103 an hour for July 2025 to June 2026, and the Netflix Dubbing Agreement reference copy lists $83 an hour. ADR on US film and TV is paid under the main TV/Theatrical contract, where the day performer minimum is $1,283 from 1 July 2026. Non-union rates vary widely.

Can you do ADR from a home studio?

Yes, and it has been common since 2020. You need a quiet, acoustically dead room, picture playback in sync with your headphones, a DAW, ideally a shotgun mic to match production sound, a remote link such as Source-Connect, and a local backup recording. The studio's ADR mixer usually drives the session and the director listens in remotely.

How do I get into dubbing with no experience?

Train the sync skill first with daily drills against muted picture, take acting and improv classes, and attend ADR or looping workshops where coordinators teach. Then build a short demo showing you matching picture, and contact dubbing studios in your language directly. Loop group work is often a first booking, because it rewards improvisation and range more than a long credit list.

What is walla in film sound?

Walla is the background murmur of a crowd, recorded after the shoot by a loop group. Modern walla is improvised, researched dialogue rather than the nonsense word that gave it its name, and it is mixed low under the principal dialogue. Netflix's dubbing guidelines, for example, ask for group walla to be dropped by 10 dB in the premix.

Sources and Further Reading

• Jeff Hinton, ADR: A Primer for Filmmakers, Frame.io, updated April 2024: the three-beep cue, mic matching, boom and lav tracks, session workflow.
• AcademieCine.TV, La postsynchronisation, Guide international des métiers de l'audiovisuel: spotting, the rythmo band, labial placement, one or two choices per line.
• Simon Fraser University, ADR (Automatic Dialogue Replacement): film loops, streamers, punches and beeps.
• CalArts 2 Pop, ADR session prep guide: an hour of prep per minute of ADR, per-character scripts and cue tracks.
• Giselle Spiteri Miggiani, English-language dubbing: challenges and quality standards of an emerging localisation trend, Journal of Specialised Translation 36 (2021): Chaume's synchrony types, phonetic sync in close-ups, findings on Netflix English dubs.
• Netflix Partner Help Center, Dubbing Premix Recording and Delivery Guidelines and Mixing Style Guide for Dubbed Content: walla level, track layout, mic setup, sync and efforts.
• Susan Stamberg, NPR, Never seen and sometimes barely heard, loopers fill in Hollywood's soundtrack, February 2015.
• SAG-AFTRA, Dubbing Agreements, the 2021 Dubbing Agreement, the Modified Dubbing wage table and the Netflix Dubbing Agreement reference copy.
• AMPTP, 2026-30 SAG-AFTRA Agreement Wage Tables: day performer minimums from July 2025 to July 2029.
• LA Magazine, SAG-AFTRA approves AMPTP deal with expanded AI protections, June 2026.
• Deadline, SAG-AFTRA deal summary released ahead of ratification vote, November 2023: the prepaid looping day provision.
• Authors Guild, SAG-AFTRA agreement establishes important AI safeguards: consent exceptions for dubbing and post-production.
• Reuters via Yahoo News, Voice actors push back as AI threatens dubbing industry, July 2025.
• The Hollywood Reporter, Netflix's global reach sparks dubbing revolution, 2019.
• Amazon, Prime Video launches AI-aided dubbing pilot, March 2025, and Anime News Network, Addressing the Amazon AI dubs controversy.
• Broadcast, Life in an audio post house during Covid-19, and Production Expert, Remote working workflow for recording dialog, ADR and dubbing, 2020.
• Sound On Sound, Steinberg Nuendo 5 review, Producing narration and dialogue for VR and 360 video and Synchro Arts Revoice Pro review.
• Production Expert, EdiCue v4 announced and VoiceQ tested, and Avid, Cue Pro.
• Post Magazine, MPSE presents 73rd Annual Golden Reel Awards, March 2026.
• SAG-AFTRA Foundation, ADR Voiceover and Looping with Lynnanne Zager.
• Sennheiser, 50 years of the MKH 416.