MugOke Vocal Reducer

Practice karaoke with on-device vocal reduction, pitch controls, effects and synced YouTube lyrics.

Make room for your voice with MugOke Vocal Reducer. Adjust the vocals and music in supported web videos for karaoke practice, singing along, or exploring a song's vocal track. A compact digital mixer puts vocal levels, pitch and song vocal effects in one place. YOUR MUSIC, YOUR MIX • Reduce song vocals to practice singing with the instrumental track. • Adjust Vocals and Music independently from 0% to 100%. • Change song pitch from −6 to +6 semitones while keeping the tempo. • Add Reverb for hall ambience or Chorus for wider, doubled song vocals. • Enable video synchronization to help align the picture with processed audio. • Choose Sharp or Smooth picture settings, GPU saver, and optional denoise. FOLLOW THE LYRICS • Synced lyrics for YouTube and YouTube videos embedded in MugOke. • Automatically load available YouTube captions when Lyrics is enabled. • Select a caption language, using YouTube auto-translation when needed and available. • See current and next lines, with three preparation dots in the three seconds before vocals start or resume after a long gap. • Choose highlight and text colors, a dark background and timing adjustment. • Tap a timed line to jump to it, import a local LRC file, or optionally search LRCLIB. • Compact menus in English (default) or Thai. • Lyrics can run independently of AI audio processing. GET STARTED 1. Open and play a supported video, such as a YouTube video. 2. Click the extension and select Start vocal reducer. 3. Adjust your mix, pitch and effects. 4. Select Stop to return to unprocessed audio. 5. Enable Lyrics to follow available captions. Pin MugOke using Chrome's Extensions menu for quick access. ON YOUR DEVICE Audio processing runs locally using the bundled AI model and WebGPU. Audio and video are not uploaded to a server. Mixer preferences are saved locally. The extension does not include ads or analytics and requires no extension account or subscription. BEFORE YOU START A WebGPU-compatible GPU and browser are required for AI audio processing. Compatible integrated GPUs may also work. Startup time, audio delay and separation quality vary by device and recording. Vocal reduction is experimental and may leave audible artifacts; protected or inaccessible media may not work. Captions and translations depend on YouTube and the video. Some videos have no usable captions or translation. Line-based highlighting is approximate, not guaranteed word-level timing. Lyrics cannot be generated without a timed source. YouTube caption and translation requests go to YouTube. Optional LRCLIB search sends your typed song/artist query to LRCLIB after permission is granted. Local LRC files stay on your device. Video synchronization uses additional memory and processing. It delays the video picture, rather than separate website lyrics or scoring overlays. Reverb and Chorus affect the separated song vocals—raise Vocals to hear them. This extension does not capture microphone audio. Player integration, remote control, and duplicate-lyrics suppression require a compatible version of the MugOke website. Vocal reduction quality depends on the original mix. Captions and translations depend on their providers. Discover more karaoke at https://mugoke.com/ Privacy: https://mugoke.com/extension/privacy Support: https://mugoke.com/about