An innovative fusion of mathematics, code, and a touch of AI that transforms city skylines into unique musical compositions.
Complete process: from skyline to music in real time
Complete transformation process: image upload, contour detection, MIDI generation, and AI musical rendering
🎬 Demonstration video of the complete CurvePhonic process
A fascinating process that converts visual shapes into sonic experiences
Upload a skyline image and watch how contours are detected
We convert contours into up to 1024 segments with mathematical functions
Adjust pitch, tempo, velocity, and duration to customize your composition
Create editable MIDI and optionally render with AI using MusicGen
No ML at the core. We use edge detection, curve segmentation, and polynomial approximation to extract the mathematical essence of each skyline.
Two specialized microservices: one for visual/mathematical processing and another for musical generation and AI rendering.
Adjust all musical parameters in real time. Generate editable MIDI that you can use in any professional DAW.
Each slope, curvature, and height translates directly into velocity, modulation, and musical pitch.
Exploring new frontiers at the intersection of visual and sonic
Splines, fractals, chaotic functions for greater musical expressiveness
Mapping based on color, density, symmetry, and other visual characteristics
Platform to generate, remix, and share compositions between users
A musical collection of iconic skylines converted into unique compositions
CurvePhonic is currently in active development. We're fine-tuning the algorithms, perfecting the UI, and preparing for a public release that will let everyone transform their favorite skylines into music.
Refining mathematical curve analysis
Creating an intuitive experience
Scaling for public access
Be among the first to experience CurvePhonic when it goes public
I'm open to feedback, ideas, and potential collaborations. If CurvePhonic sparked your curiosity, let's talk.