Voxel is extracted for regional timepoint

[vreuter]

The purpose of the voxel extraction is for fitting the 3D Gaussian for subpixel localization. We don't need/do this for the regional spots, so it's wasteful and confusing to extract the regional timepoint's data. This will complicate things like resetting the index for visualisation (#406), and basic properties like "the time dimension of the 4D array for a field of view's locus-spot viewing data should have length equal to the max length of any locus time list." Right now, this property needs +1 added to account for the regional timepoint itself. A quick check on repeat 2 data from experiment 6011's 30th FOV reveals that, as expected, the greatest index of any voxel that's not all zeros is equal to the length of the longest locus times list, not equal to that value +1 (indicating the wastefulness).

[vreuter]

NB: this can only be fixed when the locusGrouping section of the rounds config is present, so this is only going to be dealt with for the visualisation step, not the actual extraction step, as the divergence in semantics / interpretation of the data would be very error prone and need to multiple times consider whether or not the experiment was using a locus grouping.