Chordially

demo ui
web page
testing
also testing

Chordially

Chordially Logo

Inspiration

Chordially is a play on the words "cordially" meaning warm/friendly and "chord". We strive to create a friendly and welcoming environment to learn guitar chords in.

Learning music can be intimidating and expensive. Private lessons cost $50-100 per hour, and many aspiring musicians give up before they even start. We wanted to democratize music education by creating an AI-powered tutor that provides personalized, real-time feedback at a fraction of the cost. Our goal was to make quality music instruction accessible to anyone with a smartphone or computer, removing financial and geographical barriers to learning an instrument.

What it does

Chordially is an intelligent music learning companion that uses computer vision to track your fingers on the guitar frets and provides instant feedback. The app:

Tracks finger positions on the guitar fretboard in real-time using Python and ArUco markers
Maps strings and 12 frets using a mathematical fret-spacing model for precise coordinate tracking
Provides live visual feedback and accuracy scoring for chords from a 285-entry chord database
Improves tracking stability with a smoothing algorithm to reduce hand jitter and lost markers by ~40%
Highlights correct vs. actual finger positions on-screen to guide learning
Monitors progress with real-time accuracy percentages

How we built it

We built Chordially using a modern tech stack focused on real-time guitar tracking:

Computer Vision: Python with OpenCV and ArUco markers for finger and fretboard detection
Fret Mapping: Mathematical fret-spacing model (1/17.817 ratio) for precise coordinate calculation
Chord Database: 285-entry library supporting beginner chords with live accuracy scoring
Tracking Stability: 5-frame smoothing algorithm to reduce hand jitter and lost markers by ~40%
Frontend Feedback: On-screen chord diagrams highlighting correct vs. actual finger positions
Analytics: Real-time accuracy percentages to monitor progress

The system detects finger positions across all strings and frets in real-time (<50 ms latency at 30+ FPS) and provides instant visual feedback to guide learning.

Challenges we ran into

Real-time finger tracking: Maintaining <50 ms latency while tracking all strings and frets was difficult. We optimized our algorithms and implemented a 5-frame smoothing system to reduce jitter.
Marker occlusion: Fingers sometimes blocked ArUco markers, causing lost tracking. We improved detection stability and reduced lost markers by ~40%.
Precision of fret mapping: Accurate positioning across 12 frets required a mathematical model for fret spacing (1/17.817 ratio).
User experience design: Balancing constructive feedback with encouragement was tricky. We iterated on on-screen chord diagrams and accuracy percentages to keep users motivated.
Chord database coverage: Ensuring the 285-entry database supported a wide range of beginner chords without overwhelming users required careful design.

Accomplishments we're proud of

Built a real-time guitar tutoring system with <50 ms latency at 30+ FPS
Designed a live visual feedback system that highlights correct vs. actual finger positions
Implemented a smoothing algorithm that reduced hand jitter and lost markers by ~40%
Developed a 285-entry chord database supporting beginner chords with live accuracy scoring
Created an intuitive UI that guides learning with real-time accuracy percentages
Successfully tested the system with rapid chord transitions and maintained reliable tracking

What we learned

Real-time computer vision requires careful optimization and calibration
Small UI/UX details, like highlighting finger positions, significantly impact user learning
Feedback presentation is key for keeping users motivated during practice
Mathematical modeling can improve precision in physical tracking systems
Testing with real users uncovers challenges that are not obvious in simulations

What's next for Chordially

Expanded chord and scale coverage for advanced players
Multi-instrument support, starting with ukulele or bass
Gamified practice sessions with streaks, challenges, and progress rewards
AR integration to provide visual guidance on finger positioning without markers
Offline mode for practice without a connected device
Integration with MIDI and digital guitars for enhanced accuracy
Continued refinement of real-time feedback and stability for faster and more accurate tracking