local z stacks

[nataliaorlova]

While working on z drift metric design we discovered that LIMS stores this data unaveraged, with negative pixel overflow issue (all negative pixels become saturated) and also saved as HDF5 file.

Ideally, stacks should be averaged, with negative pixels parsed correctly and saved as tiff. The latter is the least priority, but would simplify significantly manual manipulation of the data (loading to ImageJ)

Example experiment ID is 882551960, here is the path to stack

Mouse seeks does the averaging correctly in it's pipeline, so an example of the new output can be found here.

Thank you!

[danielsf]

As of this writing, the local z-stacks are produced by the class defined here

https://github.com/AllenInstitute/ophys_etl_pipelines/blob/main/src/ophys_etl/modules/mesoscope_splitting/zstack_splitter.py

[nataliaorlova]

here are the links to stacks, experiment ID 882551960
/allen/programs/braintv/production/visualbehavior/prod0/specimen_837581585/ophys_session_882386411/ophys_experiment_882551960/882551960_z_stack_local.h5

averaged stack for the same experiment: /allen/programs/braintv/workgroups/ophysdev/oPhysQC/mouse_seeks/882551960/local_z_stack/local_z_stack.tif