Sooooounds Visualization

Capture real-time audio from microphone using Web Audio API. The system continuously analyzes the incoming sound stream.

Extract audio features: volume, pitch, energy, and spectral centroid. Each characteristic maps to different visual effects.

User uploads an image that serves as the base for transformation. GPT-4o vision analyzes the image content and style.

GPT-4o generates contextual prompts based on image analysis and current audio features, maintaining the essence of the original.

DALL-E 3 creates new images based on smart prompts, ensuring audio-reactive transformations relate to the original upload.

Continuous transformations every 8 seconds in live mode, with additional real-time post-processing effects using Sharp.

🎵 Audio-Reactive Effects

Volume → Brightness

Louder sounds increase image brightness and intensity, creating dynamic lighting effects that pulse with the music.

Energy → Saturation & Blur

High-energy audio boosts color saturation and adds motion blur effects, making images more vibrant and dynamic.

Pitch → Color Tinting

High pitches add blue tints, mid-range creates purple hues, and low frequencies bring warm colors to the image.

Centroid → Hue Shifting

Spectral centroid controls hue shifting and color palette changes, creating smooth color transitions.