Table read

Read your script aloud

Assign a voice to each character and listen to the script. System voices and natural voices run on your device.

Starting a read Link to Starting a read#

  1. Put the cursor at the start position.
  2. Press ⌘⇧A or select View → Table Read.
  3. Press Play to read from there, or This scene to read only the current scene.

A green margin mark identifies the spoken line. The page scrolls to show that line. Pause keeps the position. Stop or Escape ends playback.

Open Voices for casting, speed, and Skip directions. Press ⌘⇧A again or × to close the bar. The command palette also offers Read Aloud From Here and Stop Reading.

You can edit during playback. Playback includes changes after the current line. Changes to the current line do not affect playback of that line.

Start, pause, or limit the read to the current scene.

Two sets of voices Link to Two sets of voices#

System voices use the speech voices installed on your device. Screencake prefers Enhanced and Premium voices on Mac.

  • Add Mac voices: System Settings → Accessibility → Spoken Content → System Voice → Manage Voices.
  • Add Windows voices: Settings → Time & language → Speech → Manage voices.

Natural voices add twenty American and eight British voices. Enable them in the Voices panel to download approximately 365MB. You can write during the download. If the download stops, select Retry to continue.

Use natural voices on a computer with Chrome, Edge, Safari 26, or the desktop app. iPhone and iPad use system voices. Very short lines can have less natural speech patterns than full sentences.

To remove the files from your device, select Remove download in the Voices panel.

Casting Link to Casting#

  • Automatic casting: frequent speakers get different voices first, until all available voices have assignments.
  • Your cast: select voices in the Cast list. This device saves selections by character name across scripts.
  • Narrator: reads headings, action, and parentheticals. The default is your system voice, or Heart for natural voices.
  • Auto: returns a character to automatic casting. With only one usable voice, character names are read before dialogue.

A character's gender in the report guides natural-voice casting. It does not affect system voices. Age and description are saved in the profile but do not change the voice.

Select an engine, narrator, and voice for each character.

What gets read Link to What gets read#

  • Dialogue: each character's lines in their voice. Dual dialogue reads left, then right.
  • Headings: read as words, so INT. becomes "Interior".
  • Directions: action and parentheticals use the narrator. Skip directions leaves only headings and dialogue.
  • Omitted scenes: read as "Omitted", without the body. Summaries, notes, and outline entries are skipped.

Set playback speed between half and two times the usual speed.

Storyboard frames follow the SHOT being read in a panel at the bottom right, then clear at the next scene. Toggle Show storyboard in Voices to hide them. This works offline and in Local Mode. Encrypted scripts do not have storyboards.

Reading

The page follows the spoken line automatically.

Privacy Link to Privacy#

Table read processes your script on your device, not on a speech server. Voice choices and playback speed are stored on this device too.

Natural voices use Kokoro-82M by hexgrad (Apache-2.0). Pronunciation uses the CMU Pronouncing Dictionary (BSD) and phonemize (MIT). Playback uses ONNX Runtime (MIT).