NOVABERDocs

Loading search index…

    Documentation menu

    App reference

    Voice Studio reference: text to speech and role exports

    26.11.1004-beta

    Novaber · Content reviewed

    On this page

    Turn text, TXT, or SRT scripts into AI voiceovers on your computer. Organize narration and character dialogue with standard voices, authorized voice cloning, line retakes, pronunciation corrections, and WAV, MP3, subtitle, or per-role track exports for videos, courses, and audio projects.

    Capabilities and input / output

    • Standard voices, licensed voice cloning, and text-designed voices

    • Batch TXT/SRT import, multi-role settings, and reusable voice templates

    • Per-line retakes, A/B selection, pronunciation dictionaries, and version rollback

    • WAV/MP3, subtitles, per-role stems, and complete delivery bundles

    Workflow

    1. Import

    Import dialogue in batches while preserving subtitle timing

    2. Voices

    Choose or reuse authorized voices for narration and roles

    3. Produce

    Preview and redo lines, correct pronunciation, and adopt the best take

    4. Deliver

    Export WAV, MP3, SRT, per-role stems, and a complete delivery bundle

    Requirements

    • GPU: NVIDIA GPU recommended
    • VRAM: 4 GB standard voices; 8 GB recommended for clone/design
    • Images and models: Shared image layers use about 7 GB; the three models require about 9.6 GB more

    Review your first result

    These checks are suggestions based on published inputs and outputs. Use a small sample to verify processing, preview and export together; no measured accuracy or speed is provided here.

    Use a short authorized script, confirm character-to-voice assignments and listen for pronunciation, pauses and subtitle alignment. Reopen audio, subtitles and character tracks after export. Current delivery is audio and supporting files; video dubbing, finished-video export and whole-book assembly are not provided.

    Version, limitations and license

    Current catalog version: 26.11.1004-beta · Beta.

    Before cloning a voice, confirm lawful authorization for the recording, voice identity, and script. Do not use generated audio for impersonation, fraud, infringement, or other unlawful purposes. Users are responsible for uploaded and generated content and how it is used. Third-party components remain subject to their respective licenses. This is a Beta release: start with a short, authorized script, listen to the audio, and check the exported files. Automatic video dubbing/export, whole-book assembly, and additional engines are not available. Report problems through Feedback on the Novaber website, including the app version, error message, and steps to reproduce.

    Product details and installation

    Organize the task by script and delivery

    Input or goalMain operationCheck before delivery
    TXT or typed dialogueOrganize paragraphs and assign narration/rolesComplete text, role assignment and voices
    SRT scriptPreserve and review subtitle timingText, timing and listening rhythm
    Standard voicesSelect or reuse a voice configurationPronunciation and pauses in a short sample
    Authorized cloning or voice designConfirm model, source permission and configurationActual voice matches the intended result
    Editing and mixingExport WAV, MP3, subtitles or role tracks as neededFiles, role scope and adopted versions

    Standard, clone and design routes have different equipment needs. Prepare the selected model first. SRT provides dialogue and timing organization; use the catalog description above for the current Beta's delivery scope.

    From dialogue preparation to audio delivery

    1. Assign roles and voices

    After importing TXT/SRT or entering dialogue, verify text, narration, roles and subtitle times. Select standard or authorized clone/designed voices; use templates when reusing settings. Validate a short script before a large first run.

    2. Listen and correct per line

    Check pronunciation, pauses, role changes and subtitle correspondence. Redo individual lines, compare A/B outputs and explicitly adopt a take. Check dictionary corrections by listening to actual output. After rollback, confirm the currently adopted content.

    3. Export for the intended use

    Use audio for listening/editing, subtitles for timing/text review and role tracks for mixing. Reopen WAV/MP3, subtitle and required role files to check missing lines, incorrect roles or outdated takes.

    This Beta does not provide automatic video dubbing/export, whole-book assembly or additional engines. Finish video production in the relevant editor and inspect the full result. See voice delivery questions.

    Prepare the environment

    For NVIDIA GPU workflows, check driver major >= 575 and Windows driver >= 576.57 for the project’s CUDA 12.9.1 baseline. See the driver and CUDA baseline for version comparison, unknown results and container checks. Check the selected app’s GPU model, VRAM and model requirements separately.

    Use the versioned setup skill for assisted checks.

    Sources