Loud & Wrong

Inspiration

Voice should feel powerful, not embarrassing. So, I wanted to build a game that makes people drop the "I sound weird" or "I feel shy" insecurity fast, by turning any sound (whisper, talk, scream) into a clean, competitive mechanic.

What it does

Loud & Wrong is a neon arcade “voice tennis” duel. Make sound when the ball enters your zone to return it—volume controls speed, pitch unlocks power shots. Play 2-player local or solo vs AI. First to 11 (win by 2). Match card at the end for sharing.

How I built it

Web app with real-time microphone input, FFT frequency analysis, and a 60fps canvas game loop. I added calibration + sensitivity profiles (Shy/Average/Loud) so quiet voices work reliably, plus AI reactions and difficulty levels.

Challenges I ran into

Voice calibration and mic permissions and device variability (noise floor, gain, different mics), keeping latency low, avoiding false triggers from constant noise, and making timing feel fair with clear visual hit zones.

Accomplishments I am happy with

-A reliable calibration that works across voices and devices

-Smooth, responsive gameplay with visual cues

-Solo AI mode that feels beatable (not perfect)

-Pitch-based special moves that add a real skill ceiling

-Shareable match summary card (stats + highlights)

What I learned

Input calibration matters more than “cool features.” Clear feedback beats complexity. If a mic app works on the first try for every judge device, one can win points immediately.

What’s next for Loud & Wrong

Online multiplayer rooms, more modes (Whisper-only tournaments, boss AIs), richer analytics (reaction time, consistency), accessibility presets, and a “party pack” with rotating challenges and leaderboards.

Built With

• 60fps
• analysernode
• berry
• css
• fft
• getusermedia
• html5
• javascript
• localstorage
• requestanimationframe