Meeting overview
-
Dipole modeling update
- Model overcompensates and under compensates
- We are reaching the limits of the B-train on flat bottom
- We cannot “flatten” the field by flattening the machine field at plateaus, since field is drifting.
-
Denoising update
- Do we keep the 50Hz ripples in the current
- CAE denoising is powerful, but tends to introduce new artifacts which are not physical. Perhaps not useful for us - but maybe for others
-
Dipole measurements update
- Current measurements only accurate to 2e-1. Can we fix this?
- Since real current is not programmed current, inference with programmed current will give a lower predicted field than with real - and likely introduce artifacts like ripples in the output