Skip to main content

Voice Casting

Voice Casting is how you assign and manage the voices of your digital actors in ACT3 AI. Every digital actor has a voice profile that determines how they sound — their language, accent, pitch, pace, and emotional register. You can use built-in AI-generated voices, import a reference recording, or request recordings from real voice actors using the built-in request workflow.

Voice Profiles​

A voice profile is attached to each digital actor and applies to all text-to-speech (TTS) lines generated for that actor in every scene. A consistent voice profile means the same actor sounds the same throughout the production — in dialogue, narration, and reaction lines — unless you specifically override it for a scene.

Voice profiles store:

  • Language and accent
  • Pitch and tone characteristics
  • Speaking pace
  • Emotional register (neutral, warm, cold, intense)
  • Any reference audio recordings used to shape the voice

Selecting a Voice​

Open a digital actor's profile in the Actor Library and go to the Voice tab to configure the voice profile.

  1. Open the Actor Library from the sidebar
  2. Open the digital actor you want to configure
  3. Click the Voice tab
  4. Browse the built-in AI voice options — they are organized by language, accent, gender, and age range
  5. Click Preview on any voice to hear it read sample lines
  6. Use the filter controls to narrow by: language, accent (regional, neutral, British, Southern US, and many others), gender expression, and age range (child, young adult, middle-aged, senior)
  7. Select the voice that fits the character's personality, role, and the tone of the production
  8. Optionally upload a reference audio recording — a short clip of a voice you want to match. ACT3 AI uses the reference to fine-tune the selected AI voice toward that sound.
  9. Click Save Voice Profile

The selected voice now applies to all TTS lines generated for this actor. You can return to this tab at any time to update the voice.

What to Listen For When Choosing a Voice​

  • Pace — Does the voice feel unhurried or driven? Match it to the character's energy.
  • Register — Higher voices often read as younger or more anxious. Lower voices often read as authoritative or calm.
  • Warmth vs. edge — Some voices have a natural warmth that works for protagonists and allies. Others carry a natural edge that suits antagonists or authority figures.
  • Accent consistency — If your production has characters from specific regions or backgrounds, match the accent to the character's origin rather than defaulting to neutral.

Audition several voices reading the same line before committing. The difference between a good voice and the right voice often becomes clear when you hear the character's actual dialogue.

Voice Actor Request Workflow​

Non-voice team members — directors, writers, editors — can request real voice recordings from voice actors without leaving the platform. This workflow coordinates recording tasks, tracks progress, and integrates finished recordings directly into the timeline. See Collaboration for how team roles apply here.

Submitting a Recording Request​

  1. Open a scene or shot in the Editor
  2. Go to the Cast panel on the right side
  3. Find the character you need a recording for
  4. Click Request Voice Recording
  5. Confirm or write the line text
  6. Add context notes for the voice actor: scene description, emotional direction, character backstory notes, and any specific delivery guidance
  7. Set a deadline if the production schedule requires it
  8. Click Submit Request

A task is created in the Voice Tasks inbox. The task is visible to all team members assigned the voice actor role in your Organization.

The Voice Tasks Inbox​

Voice actors in your Organization see their assigned tasks in the Voice Tasks inbox, accessible from the left-hand navigation.

Each task shows:

  • Character name
  • Line text (what needs to be recorded)
  • Scene context — a brief description of the scene and the moment in the story
  • Emotional and delivery direction
  • Deadline
  • Status (Pending, In Progress, Submitted, Approved, Re-take Requested)

Voice actors can see all tasks assigned to them and prioritize their recording work from this view.

Recording and Submitting​

Voice actors record their lines in two ways:

In-Browser Recording:

  1. Open the task in the Voice Tasks inbox
  2. Click Record
  3. Record the line using the browser's microphone
  4. Review the recording by playing it back
  5. Re-record if needed
  6. Click Submit when satisfied

File Upload:

  1. Open the task
  2. Click Upload Recording
  3. Select an audio file from your device (WAV or MP3)
  4. Preview the upload to confirm it is the correct file
  5. Click Submit

Submitted recordings appear immediately in the Director's review queue.

Reviewing and Approving Recordings​

Directors and Admins review submitted recordings from the Voice Tasks panel.

  1. Open the Voice Tasks panel from the left-hand navigation
  2. Find tasks with Submitted status
  3. Play the recording in context — the review player shows the recording synchronized with the scene timeline
  4. Click Approve to accept the recording. It replaces the TTS placeholder in the timeline.
  5. Click Request Re-take to send it back to the voice actor

When you request a re-take, you can add notes explaining what needs to change — delivery, pacing, emotional intensity, or a specific word emphasis. The voice actor receives a notification and sees your notes in the task.

An approved recording replaces the TTS-generated audio for that line across all shots in the scene where that line appears.

Voice Library​

Approved recordings are stored in the Voice Library — a section of the Asset Library that holds finalized voice recordings for reuse.

Once a recording is approved, it is available to use in other scenes or episodes without re-requesting it from the voice actor. This is useful for:

  • Recurring catchphrases or signature lines
  • Episode intros and outros
  • Narration lines that repeat across a series

Access the Voice Library from Assets → Voice Library in the sidebar.

Tips for Voice Direction​

Give the voice actor the scene context, not just the line. "Say this line cheerfully" is much less useful than "This character just discovered something they have been searching for for twenty years — the line lands at a moment of pure relief mixed with disbelief."

Specify what the character is hiding. If a character is trying not to show fear, telling the voice actor to play "controlled with a hint of tremor" is more actionable than "nervous but brave."

Reference a comparable tone rather than a comparable voice. Rather than saying "sound like [another actor]," describe the emotional quality: "measured and cold, like someone who has made their decision and is past the point of doubt."