Local Audio Analysis and Recognition
Where are the quiet spots in this file?
Choose a local audio file and a threshold. This tool identifies low-level spans as candidate trim points for you to review, entirely in your browser, and never edits the file itself.
Answer
Your audio file is analyzed locally and never uploaded, cached, or included in analytics.
Want to learn more? Local Audio Processing and Privacy
Understand
A threshold is a decision aid, not a claim of true digital silence: a very quiet room tone, a fade tail, or background noise can all sit below your chosen threshold without being truly silent. This tool never edits your audio; it only lists candidate spans for you to review and act on yourself.
Worked example
A file of pure digital silence is found as one long candidate span covering nearly its entire duration. A steady tone well above the threshold produces no candidate spans at all, and a very short quiet gap shorter than your minimum span setting is ignored rather than reported as a false trim point.
Method and limitations
The file is decoded locally and divided into 20 ms windows; each window's RMS level is compared against your threshold, and adjacent below-threshold windows are joined into one span. Spans shorter than your minimum duration are discarded. This is a local-level measurement only; it does not distinguish silence from very quiet program material or noise.
Choosing a threshold that fits your source
A field recording made in a quiet room might reasonably use a threshold around -60 dBFS, while a recording with audible room tone or hiss may need a less aggressive threshold to avoid flagging program material as silence. There is no universally correct setting; the right threshold depends on your specific recording's actual noise floor, which is exactly why this control is left in your hands rather than fixed.