Or is this a more serious rendering issue, where the slices are not rendered at the middle of the slice, but in between? Or is this how it is supposed to work?

@nicole let me know if you want to review the logic with me. I'd like to see this sorted out asap.

-Steve

From the slice controller/logic side, the rotate to volume plane button is calling rotateSliceToBackground on the volume node. Adding the call to the slice logic to SnapSliceOffsetToIJK shows the fiducial still appearing on the edge of the slice in the Sagital and Coronal views. Oddly, it's showing at the other edge of the slice than without the snap call.

I'm still digging into the assumptions in jumping slices and rotating the volume and converting fiducial RAS locations to display positions to try and track down what's happening where.

This might be related or not: when I try to jump slices, very often I do not see the fiducials in the axial view, it's somewhere off. Very annoying. Especially while trying to prepare figures for a paper under a deadline pressure, using analyzed data.

Thsi seems to be related: if I volume render label drawn on a single slice, and make slice viewer visible in 3d, it does not go through the middle of the rendered label. Reproducible with the Dec 25 nightly.

Hi Andrey -

The issue you reported in note 12813 sounds independent of the fiducials. Can you create a child bug with exact steps to reproduce it and assign to me? My guess is that this is a volume renderer issue but I would like to confirm.

Thanks,
Steve

Fixed in r26082 by snapping slice viewer to voxel center after rotate to volume plane (same way as done after FOV reset).

Some other issues mentioned in the discussion have been fixed since then, too (computation of slice spacing of oblique volumes, centered jump to slice).