Voice acting

• 1. Introduction

  ›

  Voice acting gives a face to the faceless. Every time a subway door opens and a recorded voice tells you to stand clear, a voice actor made that moment. Every animated character whose face you have memorized but whose body you have never seen exists, in large part, because of a performer you likely cannot name.

  The craft reaches into films, television shows, video games, documentaries, commercials, audiobooks, radio dramas, amusement rides, theater productions, puppet shows, and audio games. It encompasses singers stepping in to fill a character's singing voice when the lead performer cannot carry the tune. It includes actors who speak lines for a computer program or a radio dispatcher who never appears on screen. And it now confronts a challenge that none of its practitioners anticipated: the rise of software capable of generating and modifying human voices without any human in the recording booth.

  How did voice acting become a recognized dramatic profession? What makes a commercial voice different from a radio narrator? And why did voice actors from Steve Blum to Jennifer Hale find themselves on the front lines of a labor dispute that touched the entire entertainment industry? Those are the threads this documentary follows.

• 2. The Craft Itself

  ›

  A voice actor playing a narrator carries the entire architecture of a story on their vocal delivery. The narrator, as the source of a tale, can be a personal character or simply a non-personal voice the creator develops to guide the audience through a plot. In traditional literary forms such as novels and memoirs, narration is required. In plays, television shows, and films, it remains entirely optional, which means when a director chooses to use one, the decision carries deliberate weight.

  Character work for live-action productions differs from animation in a practical way: the voice actor may be reading the part of a computer program, a radio dispatcher, or some other figure that never appears on screen. With audio drama, the constraints loosen further. There is no lip movement to match, no original actor whose cadence must be mirrored. That freedom allows for more expressive performance.

  Producers and agencies actively seek a wide range of vocal types. Booming voices find homes in dramatic productions; younger, softer voices serve trendier markets. Regular, conversational voices that sound like no one in particular also have a place, provided they are deployed correctly. The commercial world operates by its own rules: television spots tend to use a narrow, flat inflection pattern, while radio commercials, especially local ones, push toward a wide, almost over-the-top delivery style. The distinction is practical and deliberate.

• 3. Automated Announcements

  ›

  The "mind the gap" announcement introduced on the London Underground in 1969 is one of the most-heard pieces of voice acting in the world. The voice currently behind it belongs to Emma Clarke. That single phrase, looped across decades and heard by millions, illustrates how a short recorded fragment can outlast any individual broadcast or production.

  The simplest automated systems play back short phrases when triggered. More complex systems, such as a speaking clock, reassemble speech from fragments: "minutes past", "eighteen", "p.m." The word "twelve" can serve double duty, appearing in both "Twelve O'Clock" and "Six Twelve." Voice artists record these fragments, which a computer then stitches into coherent announcements. The applications extend to on-hold phone messages and location-specific announcements at tourist attractions.

  This category of work is not glamorous. It does not attract fan followings the way anime voice work does in Japan. But it may be the most pervasive form of voice acting in everyday life, reaching people who have no interest in animation or radio drama simply by virtue of riding an elevator or calling a customer service line.

• 4. Voice Acting Around the World

  ›

  Japan maintains roughly 130 voice acting schools, a number that reflects how seriously the country treats the profession. Voice actors there often work for a specific broadcast company or talent agency and attract their own dedicated fans, people who watch a show specifically to hear a favorite performer. Many branch into music, singing opening or closing themes of the shows in which they star. Side projects extend to audio dramas set in the same fictional worlds as the original anime, and to image songs sung in character that never appear in the show itself.

  In Brazil, most theatrical films are dubbed into Portuguese, and most Brazilians prefer watching films in their native language. Voice actors there frequently also work as dubbing directors and translators. To practice the profession formally, a person must first qualify as a professional actor and then complete dedicated dubbing courses.

  In Iran, the field divides into three distinct categories: voices recorded over Persian films, voices for Iranian animations, and dubbing of foreign films. The first category dates to a period when facilities for recording dialogue simultaneously with filming were scarce. Though those limitations have largely been resolved, voice actors still sometimes substitute for on-set audio in the current era. Foreign dubbing represents the largest share of the work, with dubbing directors adapting translated dialogue to match the mouth movements and emotional register of the original performance.

  The United Kingdom holds a particular distinction: voice acting is recognized there as a specialized dramatic profession, a status that traces directly to the BBC's long history of producing radio dramas. That institutional backing gave the craft formal legitimacy it took longer to establish elsewhere.

• 5. Dub Localization and ADR

  ›

  Automated dialogue replacement, known in Britain as post-synchronization or post-sync, asks an actor to re-record their own lines after filming ends. The goal may be to improve audio quality, reflect script changes, sharpen diction, correct timing, or replace an accent that did not serve the production. The process is also called looping or a looping session.

  Dub localization is a separate discipline. There, voice actors step into roles originated by performers in another language, altering a foreign film or television series for a new audience. Voice-over translation differs from full dubbing: the original audio track remains audible in the background while a new recording is laid over it. This technique appears most often in documentaries and news reports, where translated statements from foreign-language interviewees need to be accessible to a different audience without erasing the original speaker entirely.

  The UK's broadcasting authorities imposed an unusual constraint on this work between 1988 and 1994. During that period, the voices of people linked to violence in Northern Ireland were banned from broadcast. Television producers found a workaround: voice actors dubbed over synchronized footage of the banned individuals, preserving the visual record while complying with the restriction.

• 6. AI Voices and Industry Conflict

  ›

  In 2019, AI startup Dessa created a computer-generated version of Joe Rogan's voice using thousands of hours of audio from his podcast. That same year, Google announced a solution for generating human-like speech from text, and video game developer Ubisoft used speech synthesis to give thousands of characters distinct voices in its 2020 release Watch Dogs: Legion. The speed of the shift caught the industry off guard.

  The response from working voice actors was largely negative. The 2023 SAG-AFTRA strike included negotiations with Hollywood studios over AI regulation and separate talks with video game studios about terms to protect voice actors in that field. SAG-AFTRA described a deal it reached with AI company Replica Studios as a breakthrough, arguing it gave actors more control over how their voices could be licensed and used. Prominent voice actors disagreed. Steve Blum, Joshua Seth, Veronica Taylor, and Shelby Young all pushed back on what they saw as insufficient protections in the agreement. Jennifer Hale, David Hayter, Maile Flanagan, and Ned Luke have also spoken critically about AI voice use in video games and animation more broadly.

  The concern extends beyond labor economics. Audio deepfakes have demonstrated the capacity to put fabricated words in real mouths. In October 2023, at the start of the British Labour Party's conference in Liverpool, a deepfake audio clip falsely portrayed Labour leader Keir Starmer verbally abusing staff members and criticizing the city. That same month, a deepfake of Slovak politician Michal Simecka falsely claimed to show him discussing ways to rig an upcoming election. In January 2024, voters in the New Hampshire Democratic presidential primary received phone calls carrying an AI-generated voice of U.S. President Joe Biden, designed to discourage them from casting ballots. Each of those cases used synthetic voice technology to deceive a specific audience at a specific political moment, and none required the original speaker's consent.