Development Roadmap

This roadmap outlines planned features and improvements for Amadeus, organised by priority and timeline.

Phase 1: Core Improvements (Q1 2026)

Source Separation Integration

Goal: Extract cleaner harmonic components before transcription

Implementation:

Testing showed that the system performs well on recordings with relatively sparse textures but struggles when several instruments mask one another. This motivates the introduction of a source separation stage that extracts a cleaner harmonic component before transcription.

  • Replace unused Live Detection view with stem preview and selection interface

  • Route isolated harmonic stem to chord recogniser

  • Provide stems in app view for user selection

Benefits:

  • Higher accuracy on mixed recordings

  • Better handling of complex arrangements

  • Fits naturally into current design since pipeline expects symbolic note events

Tempo Detection

Goal: Add BPM detection and beat tracking

Implementation:

  • Use librosa’s beat tracking: librosa.beat.beat_track()

  • Display BPM in analysis view

  • Align chord changes to beat grid

  • Enable quantised chord timing

Features:

  • BPM display

  • Time signature detection

  • Beat-aligned chord grid

  • Metronome sync option

Enhanced Chord Recognition

Goal: Improve chord detection algorithm

Improvements:

  • Weighted root detection based on bass notes

  • Context-aware prediction using HMM

  • Genre-specific chord templates

  • User-guided correction interface

Phase 2: User Experience (Q2 2026)

MIDI Export

Goal: Export analysis as MIDI files

Features:

  • Chord progression as MIDI

  • Detected melody as separate track

  • Tempo and time signature metadata

  • Compatible with DAWs

Collaboration Features

Goal: Enable sharing and collaboration

Features:

  • Share analysis via link

  • Collaborative annotation

  • Comments on timeline

  • Version history

Practice Mode

Goal: Interactive learning features

Features:

  • Play along with detected chords

  • Loop sections for practice

  • Slow down without pitch change

  • Chord diagram overlays

Advanced Visualisation

Goal: Richer analysis display

Features:

  • Piano roll view

  • Frequency spectrum display

  • Chord progression graph

  • Nashville number notation

Phase 3: Platform Expansion (Q3 2026)

On-Device Processing

Goal: Remove server dependency

Approach:

  • Convert models to CoreML

  • Implement on-device inference

  • Offline mode support

  • Privacy-first architecture

Challenges:

  • Model size optimisation

  • Memory management

  • Battery efficiency

  • Performance tuning

Web Application

Goal: Browser-based version

Features:

  • Progressive Web App (PWA)

  • Cross-platform compatibility

  • Cloud sync with iOS app

  • Collaborative features

Android Application

Goal: Android native app

Implementation:

  • Kotlin/Jetpack Compose UI

  • Shared server backend

  • Feature parity with iOS

  • Material Design 3

Desktop Applications

Goal: Native desktop apps

Platforms:

  • macOS (Catalyst or SwiftUI)

  • Windows (React Native or Electron)

  • Linux (Electron or Flutter)

Phase 4: Advanced Features (Q4 2026)

Real-Time Mode

Goal: Live chord detection

Requirements:

  • Low-latency processing

  • Efficient buffering

  • Noise-robust detection

  • Visual feedback optimisation

Use Cases:

  • Live performance analysis

  • Jam session support

  • Teaching applications

  • Transcription assistance

Educational Content

Goal: Integrated learning materials

Content:

  • Interactive tutorials

  • Music theory lessons

  • Ear training exercises

  • Video demonstrations

Phase 5: Professional Features (2027)

Batch Processing

Goal: Analyse multiple files

Features:

  • Queue management

  • Parallel processing

  • Bulk export

  • Playlist analysis

API Platform

Goal: Developer ecosystem

Features:

  • Public REST API

  • WebSocket streaming

  • SDKs for various languages

  • Usage analytics

Advanced Music Theory

Goal: Deeper analysis capabilities

Features:

  • Roman numeral analysis

  • Functional harmony detection

  • Voice leading analysis

  • Form analysis (verse, chorus, etc.)

Long-Term Vision

Research Initiatives

  • Custom neural networks for chord recognition

  • Unsupervised learning from large music datasets

  • Multi-modal analysis (audio + sheet music)

  • Style transfer for chord progressions

Community Features

  • User-contributed chord corrections

  • Shared chord databases

  • Community challenges

  • Educational partnerships

Accessibility

  • VoiceOver optimisation

  • Haptic feedback for chords

  • Visual impairment modes

  • Simplified interfaces

Performance Targets

By end of 2026:

  • Accuracy: >85% on common genres

  • Speed: <1 second per minute of audio

  • Platform: iOS, Android, Web

  • Languages: 5+ supported languages

Success Metrics

  • User retention rate

  • Analysis accuracy scores

  • Processing speed benchmarks

  • User satisfaction ratings

  • Community engagement levels

Technical Debt

Items to address:

  • Comprehensive test coverage

  • Performance profiling

  • Code documentation

  • Security audit

  • Accessibility audit

This roadmap is subject to changes. Based on user feedback, technical feasibility, and resource availability.