An intelligent, voice-controlled music player and conversational AI studio.
- Broad Phonetic Wake Detection: Say "Hey Veda" or "Veda" in any tone, dialect, whisper, or deep low-pitch voice.
- Hardware AGC (Auto Gain Control): Leverages browser hardware gain adjustments to boost low frequencies and ensure quiet commands are transcribed accurately.
- Instant Voice Barge-In (< 15ms): Saying "Hey Veda" in the middle of a song instantly pauses YouTube playback, rings the acoustic wake chime, and transitions smoothly into listening mode.
- Concise Spoken Responses: Veda delivers warm, broadcast-quality spoken announcements powered by Edge Neural TTS (
en-US-AvaNeural) with WebSpeech API resilience.
- Edge-to-Edge Stage Illumination: Four corner ambient washes (
washTL,washTR,washBL,washBR), central luminescence orbs, and sweeping laser beams. - Dynamic Cover Art Color Extraction (
extractCoverArtColors): Automatically analyzes active track artwork (track.thumbnail) and adapts ambient light gradients to match each song's unique color palette live. - Audio-Reactive Beat Modulation: Real-time frequency analysis dynamically pulses orbs, beams, and washes in sync with live bass, mids, and treble.
- Crystal Clear Aesthetic: High-clarity radial diffusion designed to eliminate foggy center smudges.
- Navbar Theme Switcher: Instant toggle between crisp Pure White theme and obsidian Dark Mode (
#070b14) withlocalStoragepersistence. - Bounded Scrolling Marquee: Center ticker bounded by vertical delimiter lines displaying command suggestions.
- Dynamic Stage View Swap:
- Paused State: Veda Avatar (clickable for manual wake-up).
- Playing State: 7 rounded audio waveform bars bouncing rhythmically to live audio.
- Player Deck: Bounded track metadata, seek bar, time labels, and playback controls.
- Slide-out lyrics panel powered by LRCLIB with line-by-line synchronized scrolling.
- Click any lyric line to seek directly to that timestamp in the track.
- Dynamic background color matched to the current track's extracted cover art palette.
- Node.js (version 18 or higher recommended)
- A modern web browser with microphone access (Chrome, Edge, Brave, Firefox)
git clone https://github.com/Ish-xo/Veda.git
cd Vedanpm installnpm start(or npm run dev for auto-reloading)
Open http://localhost:3000 in your browser. Click Enter on the intro screen to unlock the studio.
| Command | Action |
|---|---|
"Hey Veda, play Starboy" |
Plays the requested song with concise spoken intro |
"Veda, play Japanese City Pop" |
Searches and streams the genre |
"Play something" / "Play music" |
Streams a trending song |
"Pause" / "Stop music" |
Pauses current playback |
"Resume" / "Unpause" |
Resumes playback |
"Next song" / "Skip" |
Skips to the next track |
"Forward 10 seconds" |
Seeks forward by 10s |
"Rewind 10 seconds" |
Seeks backward by 10s |
"Skip intro" |
Seeks ahead by 15s |
Space: Play / Pause playbackArrow Right: Seek forward 10 secondsArrow Left: Seek rewind 10 secondsKey L: Toggle Spotify-style Synced Lyrics drawerKey V: Trigger manual voice wake-upEscape: Close open drawers and panels
├── public/
│ ├── index.html # Main single-screen application markup & theme loader
│ ├── css/
│ │ └── style.css # Design system, light/dark themes, DJ lights & layout
│ ├── js/
│ │ ├── app.js # Core orchestrator, theme manager & cover art extractor
│ │ ├── audio-engine.js# Web Audio API, YouTube player & visualizer engine
│ │ └── speech-engine.js# Low-pitch STT, AGC booster & phonetic wake detector
│ └── svg/ # Brand assets & Veda avatar graphics
├── assets/
│ └── screenshots/ # Screenshots for visual showcase
├── server.js # Express backend, YouTube music search, Edge TTS & lyrics API
├── test-e2e.js # Automated end-to-end test suite (12 suites, 100% coverage)
├── package.json # Project dependencies & scripts
└── README.md # Project documentation
Run the automated test suite verifying all 12 endpoints, intents, theme switching, and visualizer elements:
node test-e2e.jsThis project is licensed under the MIT License.