Correct for compensation
Closed this issue · 0 comments
aarmey commented
We should use the provided compensation matrices to correct the channel crosstalk. The compensation matrices are added within #688. These can be renamed/reorganized as needed. I'm not sure if these are the forward matrices (e.g. one should do data * inverse(comp)
to correct) or whether they are the inverse (e.g. one should do data * comp
). However, going backwards should make the problem worse and so be pretty obvious. I do think setting this up as a matrix operation is best.