Documentation

Ensure the following baseline configuration is in place before initiating generation tasks:

• Install and launch the Vampro Voice Studio Windows companion application.
• Open Adobe Premiere Pro and access the Voice Studio extension panel (Window > Extensions > Vampro Voice Studio).
• The panel establishes a secure localhost connection to the companion service on 127.0.0.1.
• Ensure adequate disk storage on your system volume (%LOCALAPPDATA%\Vampro\VoiceStudio) for rendered stems, voice library references, and project cache.
• For 100% offline generation, confirm the English Chatterbox model pack has finished initialization.

Voice Studio provides two dedicated execution pipelines selectable directly from the top mode switch:

Build custom voice clones from source media in three steps:

1. Select Source Media: In the Studio tab, select Add Voice and choose from Timeline Clip, Bin Clip, or Upload File (supports both WAV/MP3 audio and video containers).
2. Define Sample Range: Set the in and out points to isolate a clean 5–10 second spoken passage without background music or audible overlap. Voice references must be between 1 and 60 seconds.
3. Save to Voice Library: Enter an identifiable profile name and choose Save Voice. The reference is indexed in your local library under %LOCALAPPDATA%\Vampro\VoiceStudio\voices.

Reshape a spoken performance into a target voice profile while retaining pacing, cadence, and inflection:

• Select a target voice from your Voice Library.
• Choose the source performance from a Timeline Clip, Project Bin Asset, File Upload, or live Microphone Take.
• Click Change Voice to synthesize the converted take.
• Preview the result directly in the integrated audio waveform player.
• Use "Import to Bin", "At Playhead", or "Apply to Original" to bring the file into Premiere Pro. When applying to an original timeline clip, Premiere undo history is preserved.

Text to Speech

Select Text, pick a voice profile, type or paste up to 5,000 characters of English script, and click Generate. Takes are rendered in mono 24 kHz broadcast-ready PCM WAV format.

Take Management

Generated files appear in the Preview player, job queue, and Takes catalog with duration, timestamp, and model metadata. Completed takes can be queued or deleted at any time.

Access production cleanup tools from the Repair tab:

Rate completed takes with the positive/negative feedback control. Feedback is stored strictly in your local database (studio.db). For offline voice conversion, positive ratings subtly tune the voice-strength parameter for subsequent jobs on that reference. No media or training telemetry is ever transmitted to Vampro servers.

Voice Studio stores all operational assets on your local Windows storage under:

%LOCALAPPDATA%\Vampro\VoiceStudio

Subdirectories include audio/, voices/, results/, stems/, and premiere-media/.

To remove all local Voice Studio data, terminate the companion application from the taskbar and delete the %LOCALAPPDATA%\Vampro\VoiceStudio directory.