[daip] Re: OK, I almost got it...?

[Eric Greisen]

You seem to understand.  Remaining points:

1. The reason for a CUTOFF is so that one does not apply a calibration
to data so far from that measurement that they are unrelated.  This
should mean that such data are flagged as uncalibrated because that is
what they are.  This is a user option of course - so you can
extrapolate as far as you want.

2.  The smoothing of SN is computed at the SN points - it can be
simply to replace flagged ones with smoothed ones (or replace
unflagged ones only or both).  We do not compute the smoothed at many
points although I suppose we could.  In most cases people do not do
the smoothing or use it solely to replace flagged (failed) SN points.

Eric Greisen