and do drizzle too
Hi all,
I always considered 'drizzle' to be a function of 'alignment'.
In other words the 'pre-processing' stage is broken down into the following discrete stages
1.) Acquisition of subframes - Lights, Darks, Flats, FlatDarks, BiasOffsets (the only specific requisite being, obviously, Lights)
2.) Calibration pre-processing, comprising :-
a.) Batch processing of MasterFlatDarks (optionally, obviously - and could involve BiasOffsets instead of FlatDarks)
b.) Batch processing of Flats with the MasterFlatDark (optional)
c.) Batch processing of MasterDarks (again, optional, and again could involve BiasOffsets instead of Darks)
d.) Batch processing of Lights with the MasterDark (a necessary step, even if only using a MasterBiasOffset)
e.) Batch processing of DarkCorrectedLights with the (normalised) MasterFlat (if Flats are being used)
3.) Alignment, or registration, of all CalibratedLights, choosing from the following :-
a.) Simple alignment
b.) Mosaic alignment
c.) Dither aligment (upsample, downsample, etc.)
4.) Integration, or stacking, of all AlignedCalibratedLights using some, or all, of the following :-
a.) Sum, Average, Median, Minimum or Maximum combination
b.) Additive or Multiplicative, Scaled or UnScaled Normalisation - if desired
c.) Noise Weighting of images (optional)
d.) Outlier Pixel Rejection (from simple Sigma-Clip algorithms to sophisticated Winsorized Iterative methods)
Which then gives a final image ready for the 'hard work' to start
As I see things, we already have stages 3 and 4 available in PixInsight, and step 1 is up to 'us', the 'imagers'.
So that really only leaves step 2 - and, in fact, most of step 2 is already completely available within PixInsight - just in a very time-consuming 'manual' mode.
Am I missing anything?
Cheers,
