Length of behavior files is shorter than the spike index (hc-3/ec013.15)

Hi to the kind person who can answer this message in the middle of the summer holidays.

I am trying to use the data from the repository hc-3/ec013.15. The number of video frames from the .whl file is 39985. The las index of some of the .res files is higher than 20.000.000. If I divide the last index of the .res by the sampling rate 20kHz it gives a reasonable length of a behavioral session of 17min, but the behavioral video is to short for this. I believe there is a problem here and I would need some help to verify if the sampling rate of the video is lower than the one from the electrophysiological recordings.

Thanks in advance,

Pablo.

The .whl files are described in the document for the hc-2 dataset.

https://crcns.org/files/data/hc2/crcns-hc2-data-description.pdf

It says sampling rate of the video is 39.06 Hz.  So, I think the entries in the .whl file are also at this rate.  If that is right, the duration of the video from the .whl file would be: 39985 frames /  (39.06 frames/sec) / (60 seconds/min) =  17.06 minutes.  This seems to agree with the length of the session as determined from the .res file.

Dear Jeff, 

you are absolutely right. I look at the frame rate on the original paper, but I didn't realize that for this recordings where different. As you said it is 1250/32 = 39.0625 Hz. It was my mistake.

Many thanks,

Pablo.