benjamin-heasly / gold-lab-nwb-conversions Goto Github PK

I think right now, we convert a Plexon file with nonzero plx_t_start / tStart, we'll end up with misaligned data in the final FIRA struct and/or NWB file.

When we get to the point of combining data from the original Plexon file with data from Kilosort we need to account for whatever the plx_t_start / tStart was.

In our NWB conversion, I think we can handle this as a starting_time argument to (add_kilosort_recording)[https://github.com/benjamin-heasly/gold-lab-nwb-conversions/blob/main/nwb/python/kilosort_recording.py#L64].

For (phyToFira())[https://github.com/benjamin-heasly/phy-to-fira/blob/main/matlab/phyToFira.m] maybe we can break out a similar argument and add an offset to the spike times.

Probably not a big deal, but we should keep track of this!