Precision must come from the record
If a witness says an event happened after six, preserve that wording or represent it as a time window. Do not convert it to 18:01 merely because your timeline format expects a number. A neat sequence can become misleading when approximate recollections and precise recorded times look equally certain. Make the different levels of precision visible before comparing them.
Worked example: the overlapping visits
A fictional interview places a visitor at a shop between 18:00 and 18:30. A receipt records a transaction at 18:20 somewhere else. These facts are not automatically incompatible. The interview's range may leave room for movement, or the receipt may not identify the visitor at all. First establish what each record locates and how narrowly. Only then ask whether the supported windows can overlap or follow one another.
Use a boundary question
Instead of asking for an exact arrival time that the file may never provide, ask which boundary matters to your theory. Must the person be present before a particular recorded event? Does the latest possible departure still leave an unexplained interval? In Somatry, a well-supported range can be more useful than a guessed minute because it exposes what the evidence rules out while keeping the remaining possibilities honest.