Voice fingerprinting
Build a lightweight model from short samples so the system knows who to remove.
Sports audio, your way
Mute the voice. Keep the moment.
MuteCaster is building voice-selective audio for live sports. Train it on any announcer voice you don’t want, and keep the crowd, riders, and real reactions intact.
Low-latency
Designed for live streams without lip-sync drift.
Trainable
Pick any voice you want to remove.
Cross-platform
Starting on PC, then iOS & Android.
Phase 1: Detect & Mute
Detect a target announcer in real time and mute only that voice while preserving the rest of the mix.
Phase 2: Voice-aware noise cancel
Suppress the selected voice while keeping ambience and other speakers.
Phase 3: Micro-summaries
Generate quick voiceovers of what was said in 5–10 seconds.
How it works
Three layers of intelligence, tuned for speed and clarity.
Real-time routing
Use a virtual output device: select MuteCaster as your output, then route to your speakers or stream.
Adaptive processing
Balance latency and accuracy for live broadcasts and older devices.
Roadmap
Phase-by-phase delivery, starting with a Windows desktop prototype.
Phase 1
Detect & Mute (PC)
Voice detection, live mute, adjustable thresholds.
Phase 2
Selective noise cancel
Remove one voice, keep crowd and other speakers intact.
Phase 3
Micro-summaries
Short voiceovers that recap what was said in seconds.
Built for sports fans
Skip the chatter and hear the riders, drivers, and athletes. MuteCaster is designed for racing, football, basketball, and more.
Supercross
Motorsports
Football
Basketball
Hockey
Be first to try MuteCaster
We’re assembling our early access list now. Get product updates, beta invites, and roadmap milestones.