Help & legal
How ForensiTune works, what it measures, and what it does not decide.
Legal disclaimer
ForensiTune provides analytical similarity scores and timestamped evidence to assist in assessing potential musical overlap. It does not issue legal judgments, determine liability, or replace the advice of qualified counsel.
Any decisions about infringement, substantial similarity, or damages must be made by courts or legal experts, potentially considering factors beyond what an audio signal can capture.
Methodology
- Fingerprint prefilter on robust audio hashes for fast candidate selection.
- Melody: pitch tracking, interval sequence extraction, and DTW-style alignment with transposition invariance.
- Harmony: chroma-based chord estimation and functional harmony mapping (I, IV, V, vi, etc.).
- Rhythm: onset detection, tempo estimation, and beat-synchronous pattern comparison.
- Structure: segmentation into intro/verse/chorus/bridge and alignment of section sequences.
- Risk estimation based on weighted combination of dimension scores, tunable in Settings.
Limitations & false positives
The system down-weights common chord progressions and rhythmic clichés where possible, but cannot guarantee perfect separation between generic style and genuinely copied material. High similarity scores do not automatically imply plagiarism, and low scores do not prove originality.
All computations are designed to be deterministic and reproducible given the same inputs and parameter settings, enabling auditing and independent verification.
FAQ
- Does ForensiTune listen like a human expert?
- No. It analyzes audio signals using engineered features. Its scores are a technical viewpoint, not a substitute for musical or legal judgment.
- Can I rely on a single number to decide a case?
- We recommend treating the overall score as a triage indicator only. Always inspect the timestamped evidence and read the explanations.
- Where is my audio processed?
- In this demo, all processing is simulated in your browser. In a production deployment, audio would typically be processed on a secure backend under your control.