Incorrect return shape

[andersenthomas]

I have tried running falconz on a test dataset with 63 frames. However, using "falconz -d /path/to/data -r rigid -sf 0 -rf -1" the return nifti does not contain 63 frames. 4th dimension has a size of 1 in the resulting nifti. Running without sf and rf the returned size is 46, i.e. smaller than the original data.

Is there something I have misunderstood?

[LalithShiyam]

Hi @andersenthomas. That is indeed strange and I am sorry for the delayed response. I changed my uni and all my mails and that's why I couldn't see this issue.

Is there a way you can share the nifti files? Because this shouldn't happen.

Cheers,
Lalith

[Keyn34]

@andersenthomas, could you also share a screenshot of the folder contents FALCON created?
How does your input data look like?

Running without sf and rf the returned size is 46, i.e. smaller than the original

The start frame (-sf) is determined automatically when not specified. That means the early frames of your image likely have insufficient information to perform motion correction, according to the NCC metrics between the frame and the reference frame. A later frame is chosen as the start frame in that case, and this is intended.