How to Make an Audio Drama

What you need

• An idea. One situation and one question is enough.
• A text editor. Any editor works; the script is plain text.
• A voice generator. This guide uses AudioDrama, which turns a script into a multi-voice MP3 with 16 built-in voices. New accounts get 2 free minutes.
• An audio editor (optional). Needed for sound effects, music and joining scenes. AudioDrama does not make sound effects or music.

Step 1: Pick a premise small enough to finish

Most first projects stall because the idea is too big. Narrow it to one situation, two to four characters and a question the listener wants answered: who left the voicemail? What is knocking on the hull? Why did the door open?

Limit the cast on purpose. Listeners tell characters apart only by sound, so each extra voice adds effort for them. An opening scene of two or three speakers plus a Narrator is plenty.

Step 2: Outline the beats

Write five to eight one-line beats before you write any dialogue. For the voicemail idea:

1. Hook: a phone lights up at 2:14 a.m. with a saved voicemail.
2. Complication: the voice on it is the main character's own.
3. Turn: it was left tomorrow.
4. Reveal: what the message says.
5. Ending: a decision, or a question left open.

If you cannot say what changes between the first beat and the last, you do not yet have a story, only a mood.

Step 3: Write the script for the ear

Your audience has no picture, so the script has to carry place, time and character. Four habits help:

• Anchor the listener in the first line. A Narrator line with time and place is the cheapest way to do it.
• Make every character sound different in word choice, not just in voice. One speaks in short bursts, another in long careful sentences.
• Use names. Characters addressing each other by name tells the listener who is speaking.
• Read each line aloud. If you run out of breath, the voice will too.

The genre conventions are covered in more depth on the radio play generator page, and the exact script layout is in the audio drama script format guide.

Step 4: Format the script and cast the voices

Put each turn on its own line as Name: line. A line with no name is read by the Narrator. Delete anything the tool cannot perform, such as FX: or MUSIC: lines, because they would be read out as if they were characters.

Open the generator, paste the script and look at the cast list. Every name in your script appears there with a voice. Change any voice with its dropdown, and give characters who share scenes clearly different voices. The page on text to speech with multiple voices lists all 16.

Step 5: Direct and generate

This is where AudioDrama differs from a plain reader. Add a short direction in square brackets where the delivery matters. Here is the opening of the voicemail story, ready to paste; it is 371 bytes with speaker tags, so it needs at least 25 seconds of credit.

Narrator: Tuesday, 2:14 a.m. In a dark kitchen, a phone lights up with a voicemail nobody remembers leaving.
Noor [yawning, confused]: Hello? Who is this?
Voicemail [flat, distant]: You have one saved message. Press one to hear it.
Noor [whispering]: That's... my voice.
Voicemail [slow, deadly serious]: Message left tomorrow, at six in the evening.

Practical advice for this step:

• Test short first. Every generation uses credits once your free minutes are gone. Generate the two or three most important lines before the whole scene.
• Keep your first script short. While a script generates, a hold is reserved from your balance and the unused part is returned. A short first script, around a couple of hundred words, fits comfortably within the 2 free minutes at the current hold setting.
• Edit the bracket, not the line. If a delivery misses, change one word in the direction and generate again. The page on text to speech with emotions shows what to try.
• Work scene by scene. One script is capped at 6,000 characters and 9,000 bytes per job.

Step 6: Edit and publish

Download each MP3 from your private library. If you have more than one scene, or want atmosphere, move to an audio editor and:

• join the scenes in order, with short gaps between them;
• add ambience, effects and music that you have the right to use;
• balance the volume so the narrator and characters sit at a similar level.

Then publish wherever fits: a podcast host, a video, a game or your own site. You keep ownership of the scripts you write; check the Terms for the rules, and the rules of the platform you publish on.

What AudioDrama does and does not do

• Does: turn Name: line scripts into one multi-voice MP3, with 16 built-in voices, per-line directions and a private library.
• Does not: make sound effects or music, mix audio, overlap voices, clone voices, import PDF, FDX or Fountain files, or write the story for you.
• Costs: 2 free minutes on sign-up, then $9 for 60 minutes or $29 for 240 minutes, bought once. See pricing.