Sometimes stock voices aren't enough — you want to sound like yourself, or you need a precisely tuned narrator character. FilmMovie Studio offers two paths to a custom voice: clone an existing voice from a clean recording, or design a brand-new voice from a text description (age, gender, tone, accent, pace). The result is a reusable voice_id you save on the canvas and then call in any audio workflow. In this guide we run the whole loop: from confirming consent through creating the voice to narration in /audio-studio.
What you'll need
- FilmMovie Studio with an ElevenLabs provider configured (both voice cloning and voice design are paid ElevenLabs features).
- For cloning: a clean single-speaker recording, ideally 1–5 minutes, free of music and noise (mp3/wav/m4a/aac/ogg/flac, max 10 MB).
- The speaker's consent — only clone your own voice or a voice you have explicit permission to use.
Step by step
- 1
Open the Workspace and start /voice-clone
In a project, open the Workspace and type /voice-clone into the chat box (or /voice-design if you want to design a voice from a description instead of cloning). Pick the skill from the list and send. The app takes the lead and walks you through the steps.
Film Movie StudioRenderFilm soundtrackNew projectSearch skills…AllWorkflowsQuickvideoShort DramaViral vertical mini-series, end to end.imageImage CreatorGenerate single images from a prompt.audioAudio StudioVoiceover, SFX and music tracks.copyApp BuilderMulti-screen apps with live preview.imageE-commerce DesignPro product shots for any platform.videoMusic VideoBeat-synced clips cut to your track. - 2
Confirm consent
Before it touches any audio, a confirmation gate appears: you must state that you have the right to use this voice — it's your own or you have the speaker's explicit permission. Click "Confirm and continue". If you choose "Cancel", no cloning happens.
Film Movie StudioRenderLullaby scoreNew projectFilesframe_1frame_2frame_3frame_4frame_560%frame_01.png1024²
RemixVariationsframe_02.png1024²
RemixVariationsframe_03.png1024²
RemixVariationsframe_04.png1024²
RemixVariationsframe_05.png1024²
Generating… 35%RemixVariationsTip: This is not a formality. Only clone voices you have permission for — never celebrities or public figures. A cloned voice must not be used for deception or impersonation. - 3
Provide a clean voice sample — or describe a voice
For cloning, attach a recording to your message or pick an audio card on the canvas; the app checks quality and auto-trims an over-long recording to a clean segment. For /voice-design you instead describe the voice in words — age, gender, tone, accent and pace (e.g. "a calm middle-aged narrator, deep velvety voice, slow pace").
Film Movie StudioRenderAI voiceoverNew projectFilesframe_1frame_2frame_3frame_4frame_5frame_6Drag an image here, or click to upload
reference.pngTip: For cloning, clean audio matters more than length. One speaker, no music or noise, and ideally 1–5 minutes of speech give a far better result than a noisy 20-minute file. - 4
Name and confirm the created voice
Enter a short voice name (e.g. "Narrator Max") and let the app create the voice. It returns a reusable voice_id. For cloning you can keep noise removal on; for a studio-clean recording turn it off to preserve fidelity.
Film Movie StudioRenderFilm soundtrackNew projectFilesframe_1frame_2frame_360%frame_01.png1024²RemixVariationsframe_02.png1024²RemixVariationsframe_03.png1024²RemixVariations - 5
Generate a test line to verify
The app generates a short demo line in your language and adds it as an audio card on the canvas. Listen to it. If the voice sounds wrong, have it generate a different line, or redo the creation with a cleaner or longer sample.
Film Movie StudioRenderLullaby scoreNew projectFilesframe_1frame_2frame_3frame_460%frame_01.png1024²Generating… 42%RemixVariationsframe_02.png1024²Generating… 68%RemixVariationsframe_03.png1024²Generating… 55%RemixVariationsframe_04.png1024²Generating… 80%RemixVariationsTip: If the demo language differs from the language of the original recording, the clone can sound a little less faithful — test the voice in the language you'll actually use it in. - 6
Find the saved voice card on the canvas
The voice is saved as a voice-registry card right on the canvas — it holds the voice_id, source, language and a consent note, with the demo audio linked beneath it. This card is the durable record that survives an app restart and a closed session.
Film Movie StudioRenderAI voiceoverNew projectFilesframe_1frame_2frame_3frame_4frame_5frame_01.pngRemixVariationsAdd to timelineFullscreenDeleteTip: Never keep the voice_id in chat only — the registry card on the canvas is what you remember and copy into later workflows. - 7
Use the voice for narration in /audio-studio
Start /audio-studio, paste your narration text and pass the voice_id from the registry card as the voice. The app generates the narration in your own voice and drops an audio card on the canvas. Then add it to the timeline's dialogue track and render it together with your video.
Film Movie StudioRenderFilm soundtrackNew projectFilesframe_1frame_2frame_3frame_4frame_5frame_6Drag an image here, or click to upload
reference.png
What you get
You have a custom, reusable voice — cloned from a recording or designed from a description — saved as a registry card on the canvas and verified with a test line. You now use that same voice_id in /audio-studio, audiobooks and videos: just pass it as the voice parameter when generating narration, and your projects sound consistently in your own voice.
Pro tips
- Both cloning and voice design are paid ElevenLabs features — if a call fails, the app shows the provider's original error and won't fabricate a fake voice_id.
- Designed voices (/voice-design) are experimental and work best in English, Chinese and Japanese; test them first for other languages.
- One voice_id is enough: reuse it across sessions and skills instead of re-cloning the same recording.
Try it yourself
FilmMovie Studio is a desktop app that walks you from idea to finished export. Download it and start creating.
Get FilmMovie Studio