Strategies for Flats Across Multiple Nights of Imaging

12 replies620 views
Derek Vasselin avatar
I'm usually trying to image targets across multiple nights and have noticed some slight issues with calibration frames.

I find sometimes artifacts on the imaging train move around from night to night (gusts of wind blowing stuff around, etc). Other times, I may not center the image in the exact same spot and so the dust motes and stuff come up in different parts of the image. 

This sometimes leaves artifacts on the calibrated image (see attached). 

My question:

Is there a way to manage this?

Is it possible to take flats after each night and process them with that night's batch of images, then after, take the 3-5 calibrated images and stack those? Seems there must be a better way?

For reference, I currently take flats following the first night of imaging. I use a light panel and usually take 30-40. Then any lights I get over the next few nights will get stacked all at once with the calibration frames.

I'm pretty sure it's not the flats themselves that's the issue but rather the image alignment and artifacts moving, since this doesn't happen often and I've seen these dust motes on other images pre-processing.
Well Written Insightful Respectful Engaging
Christian Bennich avatar
I do the same, do flats on the first night and at any subsequent night, IF I made any changes in the imagetrain. 

I have never seen artifacts like what you show move around.
andrea tasselli avatar
With DSLRs you take flats every single imaging section regardless whether you take down the image train or not. They do move because of magnetic levitation. Doesn't usually happens with astro-camera but it isn't impossible (as it happened to myself).
Byron Miller avatar
PixInsight WBPP can do grouping so it will associate flats with respective nights.
Xavier Ducrohet avatar
Derek Vasselin:
Other times, I may not center the image in the exact same spot and so the dust motes and stuff come up in different parts of the image.


Flats correct imperfections in your optical train. It does not matter where you point your telescope.

I have rotator, so I take flats every night (or after every target if I do 2+ in the same night). This is done automatically because I have an automated flat cover.

As Byron mentioned, you can organize your nightly files in different folder and have WBPP automatically handle it with keywords. See Adam Block's excellent video here: https://www.youtube.com/watch?v=J6tif1v8hmY&list=PLAzMa9eIVQkBmzPneF8hCgqwkrxsk7CDn&index=16 (and the next one in that playlist)
Helpful
Tom Boyd avatar
As I set up and take down every night, I take flats and dark flats at the beginning of each night - usually white I wait for it to get dark enough to begin collecting data. 

You didn’t mention which processing package you use, but I use PI. Using the WBPP script you can apply master flats and darks to each night of data and then stack multiple nights of data into a single stack.
Helpful
Kyle Cerniglia avatar
On a normal lunar cycle, I'll image for 2 weeks in broadband, then 2 weeks in narrow band. I haven't had an issue with needing to retake flats before swapping out the filters.
Well Written Concise
Rob avatar
Derek Vasselin:
Is it possible to take flats after each night and process them with that night's batch of images, then after, take the 3-5 calibrated images and stack those? Seems there must be a better way?


Uncheck the "Integration" box and run it on each night's data, matching flats to the lights.  Then stack all of the calibration files (in the calibrated folder) with the ImageIntegration process.  This is helpful as well if you collect more data on the target (year after year?), or after making changes to the image train, or cleaning anything.
Helpful
Derek Vasselin avatar
Tom Boyd:
As I set up and take down every night, I take flats and dark flats at the beginning of each night - usually white I wait for it to get dark enough to begin collecting data. 

You didn’t mention which processing package you use, but I use PI. Using the WBPP script you can apply master flats and darks to each night of data and then stack multiple nights of data into a single stack.

Yes, I forgot to mention I currently use DSS and Photoshop (I'm still new to AP). I do have plans to get PI in the near future. Knowing I can associate flats by night would fix the issue so looks like getting that is the next priority...
Christian Bennich avatar
Derek Vasselin:
Tom Boyd:
As I set up and take down every night, I take flats and dark flats at the beginning of each night - usually white I wait for it to get dark enough to begin collecting data. 

You didn’t mention which processing package you use, but I use PI. Using the WBPP script you can apply master flats and darks to each night of data and then stack multiple nights of data into a single stack.

Yes, I forgot to mention I currently use DSS and Photoshop (I'm still new to AP). I do have plans to get PI in the near future. Knowing I can associate flats by night would fix the issue so looks like getting that is the next priority...

I tried different other solutions when I jumped into the AP rabbit hole a year ago, in my perspective PI is by far the best overall piece of software you can invest in for anything processing. 
Steep learning curve - but highly rewarding!
DavesView avatar
I take flats and biasses smile after every night. PI has a very cumbersome method (file management becomes an issue) for stacking multiple sessions, so I use APP for multi-session. I prefer to leave the files on my capture drive and call them up from there. PI has you put them in a keyword folder and under that folder, N1-N2-N3 and so on. After processing, I delete all of the flats/biasses and use the masters from then on. I keep the original lights. If I only shoot one night, I always go with PI/WBPP. NOTE: I currently only shoot OSC. That is about to change.
Helpful
Rob avatar
Derek Vasselin:
Yes, I forgot to mention I currently use DSS...


Ahh, silly of me to assume!  You can group sessions in DSS also, and it might actually be a little more intuitive.  Be sure you're on the latest DSS also:  https://github.com/deepskystacker/DSS/releases


Edit:  I believe it will stack all groups together after calibration, so it isn't something you'd want to do for LRGB/SHO stuff.  I might be wrong there, as it's been a while since I've used this feature.
Helpful Respectful
Jerry Gerber avatar
When I do multiple nights of imaging (which is very difficult from the exceedingly foggy and windy San Francisco skies) I always take flats in the morning following the previous night's session.  If I image for 3 nights, I take 3 sets of flats.   It works well for me.
Well Written Concise
Related discussions
What Could Be Causing This Random Noise All Of A Sudden?
Recently when processing data on several deep sky targets I've been imaging within the course of the last month, I've encountered seemingly unexplainable phenomenon with the stacked data, I am noticing very high levels of noise AFTER stretchi...
Nov 29, 2022
Both posts describe astrophotographers experiencing unexpected image quality problems that have developed in their otherwise established imaging workflows.