quick local normalization question......

[sharpie78]

Hi everybody. Just a quick one about when to do local normalization with data sets from different nights...........

Should I be locally normalizing each nights subs to a reference frame from that night ?

or ........

Should I be locally normalizing all subs from a data set of different nights to the best reference frame of them all ?

Thanks in advance

nb. At the moment I am saving each nights weighted, registered and locally normalized subs, then integrating them all once I've got 5-10 hrs of data.

[John_Gill]

Hi,

With my workflow I do the following for each session:
Image Calibration
Cosmetic Correction
Debayer (if using OSC)

 *** All Images for each session are now clean ***

Star Alignment all images from each session
Local Normalization for all images
Image Integration for everything.

I hope this helps
John

[airscottdenning]

Is this preferred to checking "normalization" in Image Integration? What's the advantage either way?

[sharpie78]

Whether it is preferred depends on the quality of the data you're processing I guess. Just checking "normalization" in Image Integration works well with some of my data. If my neighbours are being particulary awkward and have all their lights blaring I find using local normalization helps.

[sharpie78]

Thanks John. I was fully calibrating, weighting, normalizing etc each nights data set seperately then integrating everything once I had 2 or 3 nights worth. It's worked ok so far but I'll try your method too.

For clarity......are you weighting your subs with subframeselector ? If so, before you align the subs, do you analyse and weight all the "clean" subs from all sessions together ?

[John_Gill]

Hi,

My logic is something like this: Over 5 separate imaging nights of 20 images per night, if you pre-process each nights images up to and including "ImageIntegration" you end-up with 5 images that you register against each other to create a final image.  If you collect all the images up to the "clean" section and then do a "SubframeSelector", "StarAlignment" and "LocalNormalization" you will then do an "ImageIntegration" with 100 images.

With "SubframeSelector" I look for the lowest "Eccentricity" and "FWHM" and highest "SNR" and then make a decision.  Some people use "complex formulas" to weight each image

space is not black
John