You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We noticed that the Epoch timestamps in the SAXSWAXS data appear to be offset from their expected values.
I believe the issue may be that the workflow is adding the per-datapoint Epoch value to the start time of the individual scan. However, these Epoch values should instead be referenced to the #E value in the SPEC data file.
In other words, the timestamp for each datapoint should be calculated by adding the recorded Epoch offset to the #E value, rather than to the scan start time.
For SPEC data files containing multiple #E entries, each scan should use the most recent #E value. This ensures that Epoch offsets are referenced to the correct absolute timestamp, particularly when data collection is resumed in an existing SPEC data file after acquiring data in a different SPEC data file.
Activity
jpk96 commentedon Oct 19, 2026
For SPEC data files containing multiple #E entries, each scan should use the most recent #E value. This ensures that Epoch offsets are referenced to the correct absolute timestamp, particularly when data collection is resumed in an existing SPEC data file after acquiring data in a different SPEC data file.