Video export

[fabiansinz]

For all of the scans

experimental_scans = [{'animal_id':29156, 'session':11, 'scan_idx':10}, # sensorium_2023_1
             {'animal_id':29514, 'session':2, 'scan_idx':9}, # sensorium_2023_2
             {'animal_id':29234, 'session':6, 'scan_idx':9}, # sensorium_2023_3
             {'animal_id':29228, 'session':2, 'scan_idx':10}, # sensorium_2023_4
             {'animal_id':29513, 'session':3, 'scan_idx':5}, # sensorium_2023_5

create blocks for the videos and time stamps according to the format on README.md (if applicable).

[pollytur]

@fabiansinz @KonstantinWilleke
What do we want to do with stimuli like DotSequence. They are artificially created videos by Tobias lab, but there are no blank images in between frames and the exposure time (fps) is closer to typical videos (like 30 frames per sec), so they are essentially videos and I would export them as videos. Any other opinions? (and why if yes)

Same for GaborGridSequence and RdkGridSequence

PS for the track of records from slack -we do not change images/videos sizes in experanto and we put train-val-test ties split as a separate channel