It seem that the FISH signal is not well centered in some of the spot images for tracing. This could be possible, it seems, for timepoints other than the one in which the regional spot was originally detected, due to some drift differences. However, it seems that for the timepoint of the image in which the regional barcode was detected, the pixel values should match exactly those from the original image, with the signal centered.
Seems to be in cases of bimodal spots. Downsampling for spot detection helps. See #226 for follow-up. Will be further addressable with visualisation of spots. See #192 / #63
It seem that the FISH signal is not well centered in some of the spot images for tracing. This could be possible, it seems, for timepoints other than the one in which the regional spot was originally detected, due to some drift differences. However, it seems that for the timepoint of the image in which the regional barcode was detected, the pixel values should match exactly those from the original image, with the signal centered.
Noticed by @TLSteinacker