Where in the night did that possible pause happen?

Go from the night to the moment.

See the whole-night pattern.

Find the stretches you want to inspect without treating one nightly count as the whole story.

Choose a marked minute.

Return to that part of the recording and see the model-estimated timing, duration and acoustic type.

Listen in context.

Open the available sound and follow the marked waveform. A count becomes a specific recording you can examine.

What every line in the timeline gives you

For each night, the app lists candidate events produced by its acoustic model:

• Time — model-estimated start and end.
• Duration — estimated span of the candidate event.
• Type — the acoustic label assigned by the production model.
• Audio — an available local clip you can inspect.

A density visualization shows whether model output was spread across the session or concentrated in a particular period. Product color bands are visual navigation aids, not clinical AHI severity categories.

A record you can navigate, built on our acoustic research.

The timeline brings our audio model’s marks together with their place in your recording. Our published PSG-paired methods and task results explain the model behind those marks; the available waveform and audio let you review your own example.

Times, durations and labels are model estimates. The timeline does not measure sleep stages, oxygen or clinical severity. Published window-classification results do not establish the accuracy of every displayed event boundary.