Mapping of channels in hybrid reconstruction

[jonschumacher]

@jusack and I just ran into an issue where we wanted to do a hybrid reconstruction of Bruker data with three rx channels and an SM with six channels of which three are feedback and should thus be excluded from the reconstruction. When using the new reconstruction interface, the frequency filter takes the channels as an argument but assumes that both measurement and SM have the same amount of channels. This should be changed to some sort of channel matching via e.g. a dictionary. This would also allow for reconstruction where channels are switched.