Mismatch in video timestamps and trial timestamps

AllenInstitute / npc_sessions

Tools for accessing and packaging data from behavior and epyhys sessions from the Mindscope Neuropixels team, in the cloud.

1 stars 1 forks source link

Mismatch in video timestamps and trial timestamps #114

Closed shailajaAkella closed 2 months ago

shailajaAkella commented 3 months ago

Sessions effected: 660023_2023-08-09 , 666986_2023-08-15, 644867_2023-02-23.

The timestamps from the trial structure were greater than the timestamps from the behavior video for these sessions.

trials= session.trials[:].stop_times.values[-1]

bjhardcastle commented 2 months ago

I think I found the cause of this:

sync and mvr were started long before the actual experiment (presumably for a pretest) but not stopped
there's a short gap in the camera frame times on sync, where MVR stops briefly and starts recording the files for the experiment:
(x: sample time on sync [s], y: eye camera frame interval [s])

I'll try adding something in npc_mvr to check for blocks of frametimes like this, and discard all but the one that starts closest to the start time of the experiment video data, or has the closest match in number of frames