Jon Cluce

MCFLIRT seems comparably confusing: [ What’s the relation between FSL’s motion parameters and the affine transformation matrix (specifically the translation)](https://neurostars.org/t/whats-the-relation-between-fsls-motion-parameters-and-the-affine-transformation-matrix-specifically-the-translation/3839)

Alex's idea: See if these different affines result in different FDJs. If not, we don't really need the affine-from-filtered-parameters besides as an intermediate between the filter and FDJ

Looks like the difference in the translation vector doesn’t matter much and neither does one-pass vs. two-pass for FD-J.  

Resolved in #1919 ([v1.8.6](https://github.com/FCP-INDI/C-PAC/releases/tag/v1.8.6))

This may actually be a design decision: > Looking at the code where [this command is generated]https://github.com/FCP-INDI/C-PAC/blob/v1.8.4/CPAC/image_utils/spatial_smoothing.py) and called https://github.com/FCP-INDI/C-PAC/blob/12c4f74a3cfeb544239e34301853b90b34d7edc1/CPAC/pipeline/engine.py#L700-L737, only derivatives (the ones with “Yes” under “To Smooth” in...

> in older versions of C-PAC (at least for v1.7.0, which is the one we have used for other pipelines) the functional to standard images were also smoothed and saved...

We could also / additionally replace all such `subprocess` calls to use https://github.com/FCP-INDI/C-PAC/blob/5b2cdeb400ed61bf5ddf052eae47c410446883f9/CPAC/utils/monitoring/custom_logging.py#L75-L128 and update that function to require list inputs instead of string inputs

Noting out-of-scope things I'm coming across in this refactor that I'd like to fix: - [ ] `inputnode.set[…] DeprecationWarning: use "HasTraits.trait_set" instead` - [ ] `MatplotlibDeprecationWarning: The register_cmap function was...

> In our defense, I don't think we tell/show the user that we were calling it "deoblique" and it was just simply the data "initialization" step. Yeah, I think the...

> then... is this https://github.com/FCP-INDI/C-PAC/blob/96db8b0b65ab1d5f55fb3b895855af34d72c17e4/CPAC/longitudinal_pipeline/longitudinal_workflow.py#L570-L629 just some sort of hacky way to use the output ingress and `CPAC.engine.gather_pipes()` to write out the results of this multiple times for each session...?...