Joshua Oluseyi

Conversational Media Production Hub

Conversational Media Production

Control lead-vocal replacement, audio processing, and video editing from one frontend screen. Supported local commands run without API keys.

Ready for local commands.

    Lead + Backing Vocal Swap Studio

    Upload individually separated audio tracks. Keep or replace the lead, the backing singers, or both. Everything in this browser editor stays on your device.

    1. Load tracks

    Need separated stems? The included optional Demucs CLI separates music from vocals. Lead-versus-backing vocals must be supplied separately or generated with an appropriate specialized separator.

    2. Route and mix

    3. Conversational vocal control

    Commands: “keep backing and replace lead”, “replace both”, “mute lead”, “swap backing from 12 to 25”, “lead gain -3”, “export wav”.

    Ready.

    4. Video soundtrack replacement

    After mixing, choose “Render WebM” to replace the audio in your video using the generated mix. Supported browsers require canvas capture and MediaRecorder.

    Use only recordings and voice identities you own or have permission to process. Neural voice identity conversion needs an independently configured model; this editor performs waveform routing and mixing.

    Lead-only swapping preserves the provided backing track, not backing vocals already embedded in a single mixed file. Browser voice dictation may require a browser service; typing works offline. Exported files remain under your control.

    Share this page by QROpen link