Dimau Voice Studio — Voices, Dialogue & Music

Dialogue workspace

Script and expressive directionsNarrator / Warm

Select a cue to edit its script. Use the inspector to shape its voice.

Voice timeline

Browser voice audition

Record a real voice

Your microphone · your performance

With microphone permission you can record your own narration, download it, or send it directly to the Video Studio timeline on this device.

Timeline sec

Audio delivery

Editable projects, speech renders and captions

Download a render kit for your PC, produce speech with your chosen engine, then import its audio pack into Video. Unchanged takes are reused and captions follow measured speech lengths.

Chatterbox / Kokoro models download their weights on first use; inference is local afterward. Browser previews are not recorded to audio files. Paid providers are never called by this web page.

Melody sketch

Pitch, phrasing and lyric syllables

Sketch notes and syllables, audition the instrument, then export MIDI for your music or licensed singing software. This preview is an instrument, not a singing voice.

Tempo BPM

Note

Beat

Beats long

Syllable

Voice Studio ready. Edit a script cue or record your voice.

Voice direction

Character / actor

Start (sec)

Target length (sec)

Direction

Render engine

Engine voice ID

Language (BCP 47)

Pace (0.5–2)

ElevenLabs voice controls

Speech model

Stability (0–1)

Similarity (v2 / Flash)

Style strength (Multilingual v2) Speaker boost (Multilingual v2)

v3 uses expressive directions and stability 0, 0.5 or 1; keep pace at 1. Multilingual v2 and Flash support pace 0.7–1.2; expressive tags are removed. These settings apply to the downloaded render kit, not the browser audition.

Preview pitch (0.5–2)Browser audition voice

Voice reference & permission

Only use your own voice or recordings for which you have explicit rights. Chatterbox reference audio stays in your downloaded local render kit.

Reference recording (optional, ≤10 MB) I am the speaker or have the speaker's explicit consent for synthetic voice generation.

No voice reference selected.

Scope and provider limitations

ElevenLabs and Murf: paid-key local render integration. Hume Octave: API ending November 2026; not linked. PlayHT: unavailable. LOVO and Descript: inspiration for editing workflows; no undocumented endpoints used. Synthesizer V: MIDI export only; no proprietary synthesizer code copied. Fiji language/accent support must be evaluated per installed model.