---
title: "Voice Acting Scripts: Original Rights-Safe Scenes with Cues"
description: "Write voice acting scripts with original, rights-safe scenes and performance cues. Copy three sample scenes, a cue key, and a self-tape checklist."
canonical: "https://www.murmurtts.com/blog/voice-acting-scripts-original-scenes"
---
[Murmur](https://www.murmurtts.com/)/[Blog](https://www.murmurtts.com/blog)/Guide

Guide

# Voice Acting Scripts: Original Rights-Safe Scenes with Cues

Write voice acting scripts with original, rights-safe scenes and performance cues. Copy three sample scenes, a cue key, and a self-tape checklist.

![](https://www.murmurtts.com/murmur-icon.webp)Murmur·October 20, 2026·11 min read[Download free (3-day Pro trial)](https://murmur-updates.tarunyadav9761.workers.dev/download/latest)[Buy $49](https://murmur-licenses.tarunyadav9761.workers.dev/checkout)

Free plan with Kokoro voices, scripts up to 1,500 characters · 3-day Pro trial, then $49 once

On this page

What makes a voice acting script rights-safeA cue key for voice acting scriptsThree original rights-safe scenes with cuesTurn the cue into audio with a local voice studioConsent, ownership, and disclosureWhen Murmur is not the right choiceRehearse and self-tape checklistFrequently asked questions

[Back to the journal](https://www.murmurtts.com/blog)

**Direct answer:** a voice acting script is the text a performer speaks, and the cues are the short directions typed beside the lines that tell the actor how to deliver them. To keep a practice or demo scene rights-safe, write the words yourself or build the scene from material that carries no copyright restriction, such as a public domain text, a CC0 work, or an original scene you author. This guide gives you a cue key, three original scenes you can adapt, and the steps to record a clean take.

Most "free voice acting scripts" online are copied film or TV scenes, which are not free to perform or distribute even for a class. A better habit is to control the rights from the start: keep a small library of original scenes and clearly licensed source material, and note the license next to each file. That way a demo reel, a class exercise, or a client read never depends on a clearance you did not get. The scenes in this article are original and written for teaching cue notation, not excerpts from existing works.

## What makes a voice acting script rights-safe

Rights-safe does not mean "old" or "free to find online." It means you can point to a specific reason the words may be used. There are only a few clean categories, and they behave differently. Sort each script you collect into one of these buckets before you rehearse it.

| Source of the words | Rights status | What you can usually do | Watch for |
| --- | --- | --- | --- |
| An original scene you wrote | You hold the copyright as the author | Perform, record, publish, and adapt it freely | Keep a dated draft so authorship is clear |
| A public domain text | No U.S. copyright restrictions | Use and adapt the underlying text in the United States | Remove any distribution license or trademark wrapper first |
| A CC0 or public domain mark work | Rights waived or expired, per the marking | Reuse without attribution in most cases | Confirm which mark is applied, since CC BY needs credit |
| A CC BY or CC BY-NC licensed work | Some rights reserved by the creator | Reuse within the license terms | NonCommercial and NoDerivatives terms change what you may do |
| A film, TV, or novel excerpt | Fully protected by copyright | Only with permission, or where a specific exception applies | There is no word count that makes an excerpt automatically safe |

Project Gutenberg is a common starting point because its catalog is largely made of texts no longer restricted by U.S. copyright law. Its [license page](https://www.gutenberg.org/policy/license.html) makes an important distinction: the underlying text and the Project Gutenberg trademark and license are separate things. If you strip the license and all references to Project Gutenberg, you are left with a text that is unrestricted in the United States. Keep that wrapper and the trademark rules apply, including limits on charging for verbatim copies.

Creative Commons covers the middle ground. Its [license overview](https://creativecommons.org/share-your-work/cclicenses/) lists six licenses that run from most to least permissive, adding NonCommercial and NoDerivatives conditions to a basic attribution requirement. A NoDerivatives license is a poor fit for a scene you intend to rewrite, and a NonCommercial license does not suit a paid client read. For practice scenes that you want to reuse without tracking credit, an original script or a CC0 text is simpler than a CC BY text.

The U.S. Copyright Office adds the caution that people often get wrong. Its [fair use FAQ](https://www.copyright.gov/help/faq/faq-fairuse.html) states that there is no legal rule permitting a specific number of words, and that only the copyright owner can authorize a new version of a work. So a "safe" rule such as "under 30 seconds" or "under 200 words" is not a legal standard. If a scene is not your original writing and not clearly licensed, assume you need permission.

## A cue key for voice acting scripts

Cues are shorthand. The exact symbols matter less than using the same ones every time so you and a collaborator read the page the same way. This key separates actor-facing cues, which shape the performance, from technical cues, which belong to the recording or edit.

| Cue type | Example notation | What it tells the performer |
| --- | --- | --- |
| Delivery or emotion | [gently], [through gritted teeth], [warm] | The intended tone or attitude for the line |
| Pace and pause | [faster], (beat), (long pause) | Where to slow down, hold, or move ahead |
| Paralinguistic sound | [sigh], [laugh], [breath] | A non-word sound that belongs in the voice track |
| Pronunciation | say "route" (root); [pro-nun-see-AY-shun] | How a word or name should sound |
| Emphasis | [stress: never], [hit "alone"] | Which word carries the weight of the line |
| Technical or production | [off mic], [overlapping], [scene change] | A note for the booth or the editor, not spoken copy |

Not every cue survives every workflow. A live actor reads all of them. A text to speech model reads only the spoken words and its own allowed tags, so a direction such as [through gritted teeth] is guidance for a human and may be spoken aloud or ignored by a machine. Some local models accept a small set of inline expression tags typed directly into the script text, which is discussed below. Keep two layers on the page: the clean spoken line, and the cue in brackets that a human uses and a machine can be told to skip.

## Three original rights-safe scenes with cues

Each scene below is original, fictional, and written for this guide. You can perform, record, and adapt them, or use them as a shape for your own writing. The cues are examples of the key above, not a required notation.

### Scene 1: The last train (dramatic two-hander)

> MAYA: [quiet, holding something back] You closed the office two hours ago. CLERK: [clipped] The office closes at six. Every night. MAYA: [slower, stepping closer] Then who answered the phone at nine? CLERK: [a beat, then flat] A recording. Same as always. MAYA: [a short laugh, no warmth] A recording didn't know my name. (beat) CLERK: [sighs, drops the act] No. It didn't.
>
> Original fictional scene, written for this guide

This scene exists to practice subtext. The lines are plain, and the meaning lives in the cues: the held-back opening, the clipped deflection, the beat before the clerk gives up the lie. When a reader performs it, the two beats do more work than any adjective, so the direction can stay small. For a demo reel, the same text can be read twice with different cue sets to show range without writing a second scene.

### Scene 2: Pawnshop monologue (character piece)

> NARRATOR: [soft, almost fond] A brass key sat in that drawer for forty years. Somebody turned it every day of their life, and then one morning they didn't. [brighter] That's the thing about objects. They don't get to decide when they're finished. (long pause) [wry] Go on, ask me what it's worth. I'll tell you the truth. It's worth exactly what the next person needs it to be.
>
> Original fictional scene, written for this guide

A monologue like this is useful for practicing a turn inside a single speech: fond, then bright, then dry. The (long pause) is a cue to protect, because the line after it lands only if the silence is long enough. If you generate this scene with text to speech, the pause has to be added as audio silence or a break, since a pause cue is not a spoken word. Writing the pause as a separate marker makes that easy to do.

### Scene 3: Documentary narration with paralinguistic cues

> NARRATOR: [warm, unhurried] For eleven years, the lighthouse keeper wrote the weather in a single ledger. (beat) [reading, gentle] Wind southwest, four. Rain by midnight. [breath] Rain by midnight. [slower] On the last page, the handwriting changes. (pause) It stops in the middle of a sentence.
>
> Original fictional scene, written for this guide

Narration cues work differently from dialogue. Here the cues manage pace and breath rather than emotion, because a documentary read needs to sound steady even when the content is heavy. The repeat of "Rain by midnight" is deliberate: it lets a performer place a [breath] between the two readings and discover where the emphasis falls. Test the scene at two speeds. A narrator who rushes the final line loses the ending.

## Turn the cue into audio with a local voice studio

Once the script is settled, **Murmur** can turn the clean spoken layer into audio on an Apple Silicon Mac running macOS 15 or later. Murmur is a local-first voice studio for text to speech, voice cloning, voice design, multi-speaker projects, and export, and it runs the models on your own machine. That means no per-character fees and no upload of your script to a cloud service, which matters for an unreleased scene or a client test. Murmur is $49 one-time, has a free plan and a 3-day Pro trial, and includes a 7-day refund policy.

Cues meet synthesis in one place: inline expression tags. Some local models in Murmur accept a small set of tags typed straight into the script text, such as reaction tags for a laugh or a sigh, or vocal events for a designed voice. Because model support varies, keep the cue in brackets while you draft, then check the model list in the app for whether that model honors inline tags before you rely on them, and remove anything the model does not support so it is not read as words. For a plain narration read, a bundled preset model is usually enough; for a specific character timbre, a cloning or design model is the better route. The [cloning, design, and conversion comparison](https://www.murmurtts.com/blog/voice-cloning-vs-voice-design-vs-conversion) explains which job each one does, and [voice design prompts](https://www.murmurtts.com/blog/voice-design-prompts-for-creators) covers how to describe a character before you record.

1. Strip the actor-facing cues down to the spoken line. Keep the cues in a separate copy of the file.
2. Split the scene by speaker if two or more characters appear, and assign one voice to each line group.
3. Generate a short test of each speaker first, then confirm the two voices are distinct before rendering the full scene.
4. Listen for a tag that was spoken as a word. If the model read [laugh] aloud, remove the tag or switch to a model that supports inline tags.
5. Export one file per speaker, then assemble the dialogue in the editor so you can retime a single line without redoing the scene.

A two-hander is a good test of the multi-speaker workflow. Assign each character its own voice, keep the line order intact, and render each part as a separate track so an interval of timing stays under your control. The [multi-voice podcast workflow](https://www.murmurtts.com/blog/multi-voice-podcast-generator-mac) covers the same assemble-and-retime loop for longer scripts with more than two speakers.

## Consent, ownership, and disclosure

Two rights questions sit beside the script. The first is the voice. If you clone a person's voice, you need that person's permission for the use you intend, and you should keep the consent with the project files. Cloning a famous voice for a private practice scene is still a permission question, not a free pass. The second is disclosure. Several platforms now expect a label when a synthetic voice is used, so plan the disclosure line before you publish rather than after. The [AI voice disclosure guide](https://www.murmurtts.com/blog/ai-voice-disclosure-youtube-creators) covers where that label belongs.

For the words themselves, keep a simple record next to each script: who wrote it, the date, and the license or a note that it is your original work. That record is what turns "I think this was fine to use" into something you can show a collaborator or a client. If you adapt a public domain text, keep the original source noted even though the underlying text carries no restriction, because the source tells you which edition or wrapper you started from.

## When Murmur is not the right choice

Murmur is a Mac app, so if you work on Windows, need live or real-time streaming speech for a call or a game, or want several people editing one cloud project at once, a cloud service will fit that workflow better. Murmur also does not do speech to text, so it cannot turn a recorded practice take into a script. If a scene needs a genuinely new emotional read that has to be discovered in the room, or a specific signed actor for a union job, record a person. Murmur is strongest when you already have the words and want private, repeatable audio you can regenerate from the same script.

## Rehearse and self-tape checklist

1. Confirm the rights record for the scene: original, public domain, or licensed, with a date and a note.
2. Read the clean spoken layer once with no cues, out loud, to find the words that are hard to say.
3. Mark one cue type at a time on a fresh copy, so you can tell which direction is actually doing the work.
4. Time the read against the intended length, then cut a repeated idea before you speed up the delivery.
5. Record or generate one take of each speaker, keep the best, and note the file next to the script line.
6. Listen back with headphones for a spoken tag, a swallowed word, or a pause that runs short, then fix the script.
7. Before publishing, add the disclosure note and keep the consent record for any cloned voice with the project.

## Frequently asked questions

## Sources

- [Project Gutenberg: the license and the public domain distinction](https://www.gutenberg.org/policy/license.html)Accessed 2026-10-14
- [Creative Commons: the six CC license types and what they allow](https://creativecommons.org/share-your-work/cclicenses/)Accessed 2026-10-14
- [U.S. Copyright Office: fair use FAQ and permission](https://www.copyright.gov/help/faq/faq-fairuse.html)Accessed 2026-10-14

## Rehearse your voice acting script in Murmur

Murmur is a local voice studio for Apple Silicon Macs running macOS 15 or later. It turns a clean script into private, repeatable audio for practice scenes and demo reads. The website edition is $49 one-time, with a free plan and a 3-day Pro trial, and a 7-day refund policy. Listen to the public samples before you buy.

[Download free (3-day Pro trial)](https://murmur-updates.tarunyadav9761.workers.dev/download/latest)[Buy Murmur · $49](https://murmur-licenses.tarunyadav9761.workers.dev/checkout)

macOS 15+ · Apple Silicon required · 7-day refund policy

Made for your Mac

## Less reading. More making.

[Download free (3-day Pro trial)](https://murmur-updates.tarunyadav9761.workers.dev/download/latest)[Buy $49](https://murmur-licenses.tarunyadav9761.workers.dev/checkout)[Explore Murmur](https://www.murmurtts.com/#product)