Audio drama and radio play writing

Write a drama the listener can follow by sound.

Let the next line change the scene.

An audio drama script plans dialogue, action and sound for a story without pictures. Separate what the audience hears from the notes that help a production team make it. The format below is a practical working convention you can adapt, not a mandatory industry standard or a promise that a text-to-speech editor will automatically cast and mix the scene.

A practical audio drama script format

Start a production draft with a title, scene number, location and a short cast list. Put the speaker’s name before each dialogue line. Keep sound cues on separate lines with a consistent label such as SOUND, followed by the event and where it starts or ends. Put delivery notes beside the relevant line only when the words alone leave an important ambiguity.

For example: “SCENE 1 — Station office, evening. CAST — Mara, a clerk; Leon, a traveler. SOUND — A key turns in a lock, then stops. MARA — That door has not opened in years.” This is a readable planning convention. Agree on the actual notation with your collaborators or production tool; ordinary labels are not universal synthesis commands.

Original two-character scene: The Unclaimed Key

The following is an original production-draft example, not an existing Honeyline recording. Setting: a station office after the last train. Mara wants to close for the night. Leon wants to identify a key left among his mother’s belongings. The sound note introduces a problem the dialogue then explains.

The Unclaimed Key — line-separated production copy
SCENE 1 — Station office,
after the last train.

SOUND — A key turns halfway in
a lock and catches.

MARA
Gently. If it breaks, we will
both be here until morning.

LEON
It is the key my mother kept.
She wrote this station’s name
on the envelope.

MARA
What was her name?

LEON
Iris Vale.

MARA
Then you have the wrong door.

LEON
You knew her?

MARA
She painted the signs. Every
platform, every waiting room.
There was one cupboard she
would never let us clear out.

LEON
I thought she had left me
something valuable.

MARA
She might have. The cupboard
is down the corridor.

SOUND — Two pairs of
footsteps, then a door opens.

LEON
Those are all her first
versions.

MARA
She said the mistakes helped
her remember how she got
there.

LEON
May I take just this one?

MARA
The board with the crooked
arrow?

LEON
When I was little, I thought
it pointed home.

MARA
Perhaps it still does.

SOUND — The key lifts from the
lock.

End scene.

Make dialogue carry a change, not a description of the screen

Give the speakers different immediate goals. Mara’s wish to close makes her opening warning natural; Leon’s question changes her attitude once she recognizes the name. Their conversation supplies the relationship and the reason to enter the cupboard. Neither character needs to announce every action for the audience.

Read the dialogue with the speaker labels covered. If you cannot tell the people apart, vary what each notices or how directly they answer. When an object matters, identify it in a line that serves the scene: “The board with the crooked arrow?” clarifies the choice while giving Mara a reason to respond.

Choose which sound cues the story actually needs

In this example, the catching key establishes a locked place, the footsteps mark a move and the final key sound closes the action. A cue should orient the listener, change the situation or support a mood. Remove a cue that competes with an important line without adding useful information.

Ask whether the event is recognizable by sound alone. A door opening is easier to identify than someone finding a particular photograph. If a reveal depends on reading a label or recognizing a face, let dialogue or narration provide that information. Decide separately how sounds will be recorded, licensed or created; a written cue is not an audio asset.

Adapt a multi-character draft for one narrator

A production script and a narration script can tell the same story in different forms. For one voice, remove the cast labels and translate essential cues into prose. Add brief attributions where the speaker would otherwise be unclear. Do not simply paste all the production notes and expect them to disappear from the spoken result.

For example: “The key caught halfway in the station lock. Mara asked Leon to turn it gently. He explained that his mother, Iris Vale, had left it in an envelope. At the name, Mara looked up. She led him to a cupboard of painted station signs. Leon chose a board with a crooked arrow. As a child, he had thought it pointed home.” This adaptation trades some performed dialogue for clear narration; it does not reproduce a full cast.

Make a next-day audio drama scene change clear in the spoken words

This standalone original example follows Niva and Pavel, two fictional adult theatre volunteers choosing an opening position for a puppet scene. The workshop has a bench. The courtyard has a low brick wall opposite its gate. Thursday evening and Friday morning are different scenes; the green bag carries the same red wooden puppet between them. The two quoted passages below form the complete spoken story. The scene headings identify the production draft for the reader and are not intended narration. No recording or performed cast exists for this exercise.

Non-spoken production heading: SCENE 1 — Workshop, Thursday evening. Complete spoken passage: “On Thursday evening, Niva set the red wooden puppet on the workshop bench. Pavel held a green cloth bag open beside her. “Tomorrow morning, can we try one opening line in the courtyard?” Niva asked. “I’ll bring this puppet in the bag,” Pavel said. Niva lowered the puppet into the bag. Pavel drew the cord closed and left the bag on the bench.”

Non-spoken production heading: SCENE 2 — Courtyard, Friday morning. Complete spoken passage: “The next morning, Friday, Pavel carried the green bag from the workshop into the courtyard. Niva was waiting by the low brick wall. Pavel set the bag on the paving beside the wall, opened it, took out the red puppet and handed it to her. Niva lifted it above the wall. “Good morning. Is this the way to the sea?” she said. Pavel walked to the courtyard gate. “I can see the puppet from here. Let’s begin beside that wall.” Niva lowered the puppet, placed it back in the green bag and pulled the cord closed. She left the bag beside the low wall. They had chosen their opening place.”

The scene change is carried by actual accepted words: “The next morning, Friday, Pavel carried the green bag from the workshop into the courtyard.” It names the elapsed time, the new place, the person moving and the bag’s journey. Niva’s position and the later gate action then establish where the two people are within the courtyard. A listener does not need to see the production headings to understand why the bag is no longer on the workshop bench.

Compare a draft that keeps the second heading but starts its spoken copy at “Niva was waiting by the low brick wall.” That version has lost the explicit next-morning connection and the bag’s movement from the workshop. Restore the bridge sentence as part of the spoken passage instead of assuming a scene number, a silence or an unnamed outdoor sound supplies it. The accepted version keeps Thursday’s packing complete, Friday’s carrying explicit and the test finished with the puppet back in its bag beside the wall. It does not claim that the whole performance has taken place.

Keep the two scene headings and any separate production notes out of a plain narration submission. If you choose the one-narrator route, supply both complete accepted spoken passages in order and review the current available narrator, controls, validation and quote in Make audio. The writing already identifies speakers and important actions; labels do not automatically create a cast or place clips on a timeline. Use an actual result for any later listening review. This example has not generated audio or established a pause length, scene-mixing action or export format.

Prepare the script for the production method you choose

For a cast performance, keep consistent scene and line labels so actors and an editor can discuss changes. Record a simple read-through to find unclear transitions before building a complicated sound plan. These are production choices you manage separately; this guide does not provide an automatic multi-actor studio.

For Honeyline’s Make audio, bring your original or authorized narrated text, select an available voice and add only supported editor cues. Review validation, sign-in and the displayed credit estimate before generation. One selected voice does not automatically assign actors to the character labels, combine separate tracks or assemble episodes. Keep the production draft alongside the narration version so you can revise either deliberately.

From an audio drama draft to an AI narration workflow

First, keep a production version with roles and sound notes. Second, prepare the exact spoken text for the narration route you choose. For The Unclaimed Key, the listener must understand whose key it is, why the cupboard matters and what changes at the end. Use the existing narrated adaptation as a model rather than assuming every label in the actor draft should be spoken.

Third, open Make audio and review the current narrator or passage voice controls and supported Sound, Background and Pause cues. Character names in ordinary text are not automatic casting commands. Fourth, check validation and the displayed estimate before generating, then review the result for speaker clarity and transitions. The preparation is useful even when a scene uses one narrator; a full cast, separate actor recordings and automatic episode assembly are different production choices.

A small audio drama generation review sheet

Write three notes beside each scene: what the listener must learn; which words convey it; and whether a sound cue adds information. For the cupboard reveal, “the board with the crooked arrow” identifies the object, while the narrated movement locates it. A door sound may support the scene, but should not be the only explanation of what Leon discovers.

After generation, compare the spoken result with those notes. If a speaker becomes unclear, revise an attribution or the wording before changing every sound cue. If a transition feels crowded, check the surrounding sentences and supported pause options. Use the current editor’s correction controls and quote; the guide does not promise a fixed cast, exact delivery or free retakes.

Open a recording copy in Final Draft and check each element

Keep the working draft before opening a separate TXT or RTF copy in Final Draft. The official file-opening instructions say to choose Script when converting either file type; this starts with the default screenplay template. The references immediately below this section describe these software steps.

To apply a different available format, use Format > Elements on Mac or Format > Elements Settings on Windows, then Apply a Template. Choose the layout agreed for your production. The checks below are a text-proofing exercise for this example, not a record of an import tested in Final Draft.

Scene target: the added heading says “SCENE 1 — Station office, after the last train.” It comes from the existing setting note. Keep it as a scene heading and keep “End scene.” as a production end note, separate from anyone’s spoken dialogue.

Character target: every MARA or LEON line identifies the following speaker. The copy contains eight Mara turns and seven Leon turns. Check each complete turn as one block, including its wrapped lines. Check all fifteen role/line pairs after import; a name interpreted as dialogue needs its element corrected, not a new spoken introduction.

Dialogue target: the first spoken words remain Mara’s “Gently. If it breaks, we will both be here until morning.” Leon’s last spoken line remains “When I was little, I thought it pointed home.” Mara then answers “Perhaps it still does.” Preserve the full sequence between them; formatting does not authorize a story rewrite.

Sound target: retain three separate SOUND cues in order: the catching key; two pairs of footsteps and a door opening; the key lifting from the lock. They remain production directions rather than a character’s dialogue. If the selected template classifies a cue as speech, change that paragraph’s element and compare its wording with the working copy again.

Official Final Draft import and formatting references

These articles support file opening and template selection. The recording copy and element checks are this guide’s original text exercise; they are not a tested Final Draft document.

Frequently asked questions

How do I make an audio drama scene transition clear without reading the heading aloud?

Put the necessary time, place and connecting action into the accepted dialogue or narration. In the two-scene example, the spoken next-morning sentence carries Pavel and the bag from the workshop to the courtyard. The production heading stays separate, while the listener can still follow the change.

How do I use an AI audio drama generator workflow?

Prepare the full spoken script, decide how speakers and actions will remain understandable, and review the actual narrator and supported cue controls. In Honeyline, Make audio starts with supplied text; character labels alone do not automatically cast or assemble a performed episode.

Is an AI radio play generator the same as a script template?

A template organizes dialogue and production notes. Audio generation needs the spoken words, supported voice choices and cues. Review which parts of the draft will be read aloud and which are notes before submitting a narrated version.

What is an audio drama script?

It plans the spoken dialogue, essential sound events and scene changes for a story told without pictures. A production draft can include cast labels and notes for performers; those notes are distinct from the words the audience should hear.

How do I format an audio drama or radio play script?

Use clear scene headings, consistent speaker names and separate sound-cue lines. The convention on this page is a practical starting point, not a universal standard. Follow the requirements of your collaborators or production tool.

How do I write sound effects in an audio drama script?

Label the sound event and note where it starts or ends relative to dialogue. Use cues that orient the listener or affect the scene. Explain a visual reveal through words if its meaning would not be clear from sound alone.

Can an audio drama have one narrator?

Yes. Adapt the dialogue with speaker attributions and narrate essential actions. The single-voice version on this page demonstrates the tradeoff: it tells the scene clearly, without claiming to recreate a performed cast.

Does Honeyline automatically cast multiple voices from a script?

Do not assume that speaker labels assign separate voices. For Make audio, prepare text for the selected available narrator and use the editor’s supported cues. Automatic actor assignment, track mixing and episode assembly are not promised by this guide.

Can I adapt the audio drama example on this page?

Yes. The Unclaimed Key is an original writing example for this guide, with a separate one-narrator adaptation. You can adapt the text into your own script. It is not a published recording or a supplied sound-effects pack.

Choose how your script will be performed.

Keep a production draft for actors and sound notes. For Honeyline’s Make audio, prepare a clear narrated version and review the actual voice and cue controls.