Roadmap

Roadmap

Moved from the README so it can evolve without commits. Roughly ordered.

Near term

• State-resolved descriptors — split multi-state sessions (machine on/off, day/night) before computing summary rows, so a night under a drone and a quiet morning stop averaging into one number.
• Catalog aggregation — a catalog command collecting all summary.json files in a corpus into one CSV for cross-session statistics.
• Tests + CI — pytest over a bundled synthetic AmbiX fixture (plane-wave + diffuse-noise generators make ground truth easy).

Analysis modules

• Cyclic-machine detector — duty-cycle finder (period, on/off spans, spectral fingerprint of the on-state) as a first-class module; currently a recipe.
• Tonal-event finder with schedule matching — bells, beeps, chimes: narrowband strike detection + test against periodic schedules (quarter-hourly clocks and the like).
• Wind/handling detector for outdoor takes (LF coherence between channels is the usual tell).
• Environmental-rhythm metric — a defensible single figure (or small vector) built from envelope power spectra at minute-to-hour scales, duty cycles, and state-alternation periods. Research question as much as feature.

Interop

• AEM export adapter — write ambiviz-compatible AEM .npz for a selected excerpt so the hand-off needs no glue code.
• BirdNET hook — species-level biophony on hi-fi windows, as an optional [ml] companion to PANNs.
• CLAP zero-shot search — free-text queries ("compressor hum", "door slam") across a corpus's cached features/excerpts.

Publishing

• PyPI (v0.1.0, trusted publishing)
• Documentation site (MkDocs → GitHub Pages)
• Zenodo DOI per release (GitHub–Zenodo integration)
• JOSS or similar software paper once tests + state-resolution land