OpenEphys Commutator and Bonsai AVI file

[Josue Regalado]

Hello,

I'm using the OpenEphys commutator and I'm automating it through Bonsai. The file it saves at the end of an experiment is just an extremely large AVI file. It makes running Minian (which usually handles several AVI files of 1000 frames each) really hard to be able to run effectively.

I was wondering if anyone has been able to adjust the Bonsai Miniscope settings to save the AVI in batches? Or if you've found a way to automate the OpenEphys commutator without also using Bonsai?

Best,

Josue

[Ed Hayter]

Hi Josue,

In the videowriter node in bonsai, have you tried changing 'FourCC' setting from DIB to FMP4? It doesn't answer your question I'm afraid, but does help reduce the file sizes massively.

Best,

Ed

[Florian Steenbergen]

Hi all,

@ Josue: I have fixed the file size by just splitting the videos afterwards into 1k frames per file (and at the same time cropping the image to the size of the GRIN-lens for that matter) both in Python. That seems to work very well with postprocessing using Caiman afterwards :)

@ Ed: I was tempted to also using the FMP4 due to size of the videos. However, apparently that appears to be a lossy video encoder. I inspected some local correlations on the FMP4 encoded videos and you could clearly see that it was different from e.g. FFV1 videos.  After seeing it I went back to FFV1, to make sure the calcium transients are as clean as possible.

I was wondering, did you try extracting the transients and comparing it to lossless compression algorithms?

Cheers,

Florian