The number of start times must equal the number of Harp times

[yonibrowning]

I get a timestamp error on one of the files I am trying to time sync. Any thoughts on what is going wrong here?

Traceback (most recent call last):
  File "C:\Users\yoni.browning\OneDrive - Allen Institute\Documents\GitHub\aind-ephys-rig-qc\src\aind_ephys_rig_qc\generate_report.py", line 361, in <module>
    generate_qc_report(directory, **parameters)
  File "C:\Users\yoni.browning\OneDrive - Allen Institute\Documents\GitHub\aind-ephys-rig-qc\src\aind_ephys_rig_qc\generate_report.py", line 92, in generate_qc_report
    align_timestamps(
  File "C:\Users\yoni.browning\OneDrive - Allen Institute\Documents\GitHub\aind-ephys-rig-qc\src\aind_ephys_rig_qc\temporal_alignment.py", line 645, in align_timestamps
    ts = align_timestamps_to_anchor_points(
  File "C:\Users\yoni.browning\Anaconda3\envs\aind-ephys-rig-qc\lib\site-packages\harp\clock.py", line 36, in align_timestamps_to_anchor_points
    raise ValueError(
ValueError: The number of start times must equal the number of Harp times

[yonibrowning]

For some clarification: its specific to just one recording (754372_2025-01-15_16-40-18), 5/6 of the others I have tested worked fine.

[jsiegle]

It's likely that one of the Harp barcodes got cut off at the beginning or end of the recording for that one stream. We need to make the alignment algorithm robust to that case.

[jsiegle]

It looks like there are about 14 seconds of this recording where the Harp barcodes were corrupted. Not sure what the cause was, but data outside that interval should be recoverable. I can manually generate aligned timestamps for this session, but will have to think more about how to deal with this case more generally.