We suggest using GREY or FFV1 for neural recordings. GREY does not do any compression of the data and, as Nate said, is likely the easiest to load up into analysis pipelines. FFV1 is a lossless compression codec and can generally cut the files sizes by around 1/3. The trade off is some video playback software doesn't decode this compression so you have to do a little more work when viewing or processing the data. You can search this forum for more info on this.
Long story short, if you are space limited, we suggest using GREY. If you want to save a bit of space, use FFV1. We don't recommend any other codecs for neural imaging data as others are lossly and often perform poorly. That said, MJPG or similar can be used with behavior recordings usually.