deal with FieldLine smart helmet sensor localization errors

[chrisNatMEG]

Smart helmet sensor localisation resulted in multiple sensors occupying same slot in a few cases when sensors were far inside their slots. In all cases they ended up in a neighbouring slot much deeper than reasonable.

Example of a correct and incorrect localisation was shared with Robert who asked to raise this issue.

[robertoostenveld]

for reference, I have copied the data to /home/common/matlab/fieldtrip/data/test/issue2369

[schoffelen]

Ah, this could also explain the fact that in one the Nijmegen recordings (I need to look this up) I remember seeing one sensor blabla-bz-1, which might be the consequence of one of the sensors being allocated erroneously (and thus a double occurrence)

[chrisNatMEG]

@schoffelen That is exactly how we noticed the problem. The "-1" is added because the name already exists. If you plot the locations in 3d you can clearly see that they occupy the same slot but at different depths.

I met with Fieldline yesterday and they have modified the localisation procedure to fix the localisation errors. Still need to confirm the results, though.