Power, recorded.
Now make sense of it.

1. Choose a file or demo. Keep your original export as a backup.
2. Check delimiter, time and PID mapping. Confirm column separation, elapsed duration, signal identity, bank and units against your source.
3. Accept with Analyze log. Apply corrections with Update preview first.
4. Inspect data quality. Review rejected rows, missing values, gaps and sample spacing before interpreting shapes.
5. Select useful signals. Start with two or three that answer your question, plus operating context.
6. Inspect the timeline. Tap or drag a chart, use its arrow keys, or move Log position. What happened here shows readings and their source times.
7. Select a region or driving phase. Zoom for a closer view; choose Phase region A below for region statistics and Inspect this phase in Run A.
8. Optional: compare Run B. Review its mapping and quality, select regions, then check the operating-condition match before What changed.
9. Optional: explore relationships. Choose signals and a pairing tolerance for the selected phase. Correlation alone cannot establish a cause.
10. Record an observation and next test. Use the local notebook below: capture evidence, review it, then save intentionally. Include the region, signals, units, conditions, timing limits and an alternative explanation.

Find patterns first. Diagnose second. Leaving or reloading clears the in-memory logs; returning starts a new analysis.