"Extract metadata from image file headers" not working with nd2 file in headless mode

[jenna-tomkinson]

Describe the bug

I am attempting to generate z-projected images from large nd2 files (>80GB). When I try running in headless mode, the CellProfiler pipeline stops at the "Metadata" section and sits without sending an error. I am assuming this is due to the "Extract metadata from image file headers" option, which requires you to press "Extract metadata" before updating (per the documentation). This is a necessary parameter in this case, as I need to be able to collect all of the FOVs within the file.

I was able to process a small example nd2 file in the GUI, so this issue, I believe, is isolated to the CLI/headless mode.

To Reproduce
To reproduce the behavior, an nd2 file needs to be used in headless mode to extract metadata and create max-projected images. Currently, I am unable share the nd2 files we are using.

Expected behavior
What should happen is that the pipeline goes through with the whole pipeline without sitting at the "Metadata" step without sending an error. It should extract the metadata from the nd2 file, then add the metadata from the file I want to add, then save the max-projected images using the metadata for file names.

Screenshots

This shows the output from CellProfiler where it stops at the Metadata module.

Desktop (please complete the following information):

• OS: Linux PopOS!
• CellProfiler Version 4.2.6
• Installed from source
• Conda env

[jenna-tomkinson]

example_cppipe.txt

Here is the example cppipe file in txt form so it can be uploaded to GitHub.

[jenna-tomkinson]

For reference, there is an issue on image.sc that looks to relate to this where the individual with a multi-page TIFF instead was having an issue extracting metadata: https://forum.image.sc/t/cellprofiler-headless-with-multipage-tif/37362/4