Add support for multi depth FOV sessions

multi-depth recordings have buggy extractions. The metadata records each scanfield*depth combination as a unique FOV, but the suite2p task then assumes each FOV needs to be repeated across the z-stack. This needs to be fixed in the suite2p task, and all multi-depth sessions need to be re-run. @k1o0 