Flats / Calibration issues

[Simon Pepper]

Hi SGL

I have been imaging for a few years now and always take my calibration frames, however the last couple times I have done this I cant see the flats are being subtracted/divided from the lights? The artefacts are present still in my master light as well as the master flat. I have not changed anything in my workflow and I cant get them to go.... A few additional details: The artefacts are present on all LRGB filters, so they must be on the sensor (I have checked cant see anything) or somewhere else in the OTA. I use the white tshirt method and AAP so set the ADU value around 27000. I was a bit lazy with this and didnt take the scope outside pointed it through a window but they look good I thought? I use WBPP in PI but OI have also tested in DSS and get the same issue so this must be an issue with how I am taking them I think? I have tried a million settings in WBPP and cant remove them. I will try taking them again but wanted to check to see if anyone had any other ideas?

 

 

Thanks

[ollypenrice]

What strikes me is that the dust shadows on the flat don't align convincingly with those on the light.  The circular shadow at lower right on the flat has a shadow equivalent on the light but the overlapping shadows just above half way up have no equivalent on the light.  A very faint pair of overlapping shadows at upper left on the flat have a dubious but more emphatic equivalent on the light.  It would be instructive to place the flat over the light in a layers program to test the shadows' alignment precisely.

Interestingly, though, the vignetting has been fixed by the flat. This leads to two contradictory conclusions:

- The flat is brightening the light in the corners.

- The flat is darkening the light where there are dark patches on the flat. These dark patches may or may not be properly aligned but if they were misaligned they would produce bright patches on the light - and they don't.

I'm going to assume that nothing at all probable is going to produce small circular bright patches on the flat (which would produce equivalent dark patches on the light) so we are left with the puzzling thought that these flats are both brightening and darkening the light in different places.

I can't explain that but I do think it risky to take flats after moving the setup. Dust moves.  Also, these dust shadows are certainly not caused by contaminants on the sensor. They are far too large and out of focus. There's a formula somewhere for calculating the distance from the chip of a particle of dust based on the size of its shadow in the image. However, your shadows are caused by something at a distance from the sensor, for sure.

Olly

[Simon Pepper]

Hi Olly

 

Thanks for this detailed answer and glad you are as confused (or almost) as I am. I have never had issues in the past with flats this is why I'm stumped. I only removed the camera after doing the flats as was concerned it was something on the sensor. however as noted nothing there and by the sounds of it you are too convinced they are further down the OTA. I have put the camera back and should be same orientation as it was to being tight so I will try running them again and see what happens here. I use PI that allows you to move a window over and to me it looked like they were matching up perfectly. As mentioned all the masters looked the same LRGB and FLATs but I assume they would as I was trying to match the ADUs vs exposure length. 

Thanks

[Simon Pepper]

I retook the flats and its made no difference. Just so I can rule most out what do flats not correct for? Could whatever this is not be dust and not being removed? Thanks

[scotty38]

11 minutes ago, Simon Pepper said:

I retook the flats and its made no difference.

You can ignore my question on the PI forum then, didn't realise it was you 🙂 🙂

[Simon Pepper]

18 minutes ago, scotty38 said:

You can ignore my question on the PI forum then, didn't realise it was you 🙂 🙂

Haha thanks Scotty yep tried that still there no idea what to do

[Simon Pepper]

I took a flat for each filter and a simple stretch applied to each this is how they look don’t understand why they are not calibrating out. 
 

not sure why the vignette changes either per filter but probably a valid reason for that…
 

L Ha S O R G B

Edited February 2, 2022 by Simon Pepper

[ollypenrice]

Vignette changing per filter could easily be the filterwheel not returning to the same precise spot. This is a very well known problem which manifests itself, above all, with flats.

Edit: Also, are you dead sure the focus hasn't moved? It rather looks as if it has.

Olly

Edited February 2, 2022 by ollypenrice

[Simon Pepper]

The focus may have changed on the NB filters as not used them for a while but certainly not on the LRGB ones as that’s what I used for M101. I keep seeing people mention darks are important as well in subtracting the flat I am using a dark library where the stack was taken through the Ha filter but fairly confident that should not impact this as it’s a dark? I’ll keep running some test here like trying to manually calibrate in PI. Not done that before so will have to check out a tutorial.

[Simon Pepper]

So I put the camera and filter wheel on another scope and got this so must be the camera sensor? Can’t be filters as it’s every filter so only leaves the camera? I will calibrate my new image on the other scope and see if they appear in the lights. Anyone got any other ideas or things I can try to fix? Thanks