singreflect

Singreflect is an open-source software library and framework for analyzing singing voices and providing reflective feedback to help users improve vocal technique. It combines signal processing, machine learning, and visualization to assess core aspects of singing such as pitch accuracy, vibrato, timbre, dynamics, and articulation. The project is designed for use by individual singers, teachers, and researchers, offering both real-time coaching capabilities and offline analysis workflows.

Its core components include a robust F0 estimator for real-time pitch tracking, formant analysis, vibrato and tremor metrics, spectral feature extraction (MFCCs, spectral flux), and timing/duration metrics. The feedback engine generates personalized practice tasks, progress indicators, and suggested exercises based on performance gaps. Visualization tools provide interactive plots of pitch trajectories, spectrogram overlays, formant tracks, and other metrics to aid interpretation. Singreflect supports a Python API for scripting, a high-performance C++ core, and pluggable interfaces for digital audio workstations (DAWs) and web dashboards. Data models allow logging of sessions, exercises, and outcomes, with options for local storage or cloud syncing and privacy controls.

Singreflect originated in a university research group focused on music information retrieval and pedagogy. The first public release appeared in 2018, and it has since evolved through community contributions, with regular releases and ongoing development. It is released under an open-source license and aims to remain cross-platform, with builds for Windows, macOS, and Linux.

It is used in music schools and online education to provide objective feedback and track improvement over time. Researchers note its modular design and extensibility as strengths, while some users caution that accurate analysis depends on high-quality input and consistent recording conditions, and that results should be interpreted by domain experts.

See also: music information retrieval, vocal pedagogy, signal processing, audio analysis.