Tutorial: inspect a first session

This example uses one camera and one sensor file. The same steps work with several cameras and many recordings.

The camera and signal data shown in the screenshots are generated examples, not recordings from an experiment.

For an in-app checklist while reviewing, open Help → Review Workflow.

1. Load the camera

Drag the video into the window. Its name appears in the left panel and its image appears in the video area. The readout over the video shows its time, timestamp-derived CFR/VFR rate, nominal rate, codec, and file size. Open Video Properties for resolution, duration, and the complete timing evidence.

2. Load the recording

Drag the sensor or tracking file into the same window. The import wizard opens for delimited text, showing the first rows as it will read them. Getting the time column, its format, and its unit right here is what keeps everything after this honest — Import sensor and recording data walks every field.

The traces appear below the video.

Use the inspector’s Filter sources and channels field to find a source or channel across the loaded recordings. Filtering only narrows the inspector list; it does not unload data or change visibility.

To put plots on a second display, choose View → Detach Plots. Close that window or choose Return to main window to restore the pane; named workspaces remember its detached position.

3. Find an event

Drag the shared time bar until you see a meaningful event. Watch the video, traces, and values in the left panel together — they are all showing the same instant.

For a copyable text view of the current channel values, sample indices, and camera frames, open the Values inspector page and choose Text view.

For a closer look, set Time span and drag the slider below the traces. Every trace keeps the same window: they sweep left to right together and restart together, because comparing them is the point.

4. Mark it

Select Flag Frame, or press M, to create an annotation at the current time. It records more than the time: for every camera loaded, it stores that camera’s exact frame index. That is what makes the exported list usable as a corrections file for retraining a pose model — see Flag frames and export.

5. Save an observation

Use Snapshot (Ctrl+E) to save the visible video and plots, or mark a range with [ and ] and export a trimmed clip or a slice of the data.

Your source files are never modified. Offsets, mappings, and annotations live in the session file beside them.

Next

If your recordings do not line up, align recordings covers offsets, drift, and fitting an alignment from TTL or frame-trigger evidence.