Dark donut in Newtonian flats? Unsure of cause or how to correct.

[pipnina]

I've been strugling for a while to get my flats to work as expected, and while this seems to be properly correcting my vignetting and dust bunnies, with the major gradient probably being my light pollution (tho it's an odd shape....) I have this ugly dark donut in the middle.

I have been taking daylight flats, with the scope pointing opposite to the sun in the sky and a thick white cloth over the opening of my 130-PDS, rubber band holding it taught with no wrinkles. Focus hasn't moved since I literally went out the morning after capture finished, and put the cloth staight on and took snaps. I target a peak of about half of full on my sernsor (around 30-34k out of 65k).

These flats were taken as the sun was setting, so maybe the gradual decrease in sky brightness is affecting them? Is it because of the secondary mirror blocking light too close to the front aperture? Some quirk of newotnians in general? given as this correction is much better than with no flats at all, can programs like PixInsight remove the complex gradient without destroying potential IFN and other faint detail currently invisible due to the gradients in my image?

This was merely a small sample stack of my nearly 6 hours of lum data, and I genuinely believe I caufght a fair bit of IFN in it but with these gradients it's impossible to draw out! : (

[ONIKKINEN]

What does a single flat frame and the flat master look like?

Did you use all calibration frames? You have over correction, which to me looks like you might have skipped darks or bias, or both. Depending on the camera and flats exposure time you may need to use darkflats to calibrate the flats too. If you used one of the newer low thermal noise cameras then bias works just fine for that (i do this, no issues).

This image does look like the skyflats had some gradient to them with one corner much better than the other, but the over correction is what bothers me most.

Edited March 27, 2022 by ONIKKINEN

[pipnina]

13 minutes ago, ONIKKINEN said:

What does a single flat frame and the flat master look like?

Did you use all calibration frames? You have over correction, which to me looks like you might have skipped darks or bias, or both. Depending on the camera and flats exposure time you may need to use darkflats to calibrate the flats too. If you used one of the newer low thermal noise cameras then bias works just fine for that (i do this, no issues).

This image does look like the skyflats had some gradient to them with one corner much better than the other, but the over correction is what bothers me most.

I'm using bias brames, and these ones match the settings of the lights (same iso, same exposure etc)

I use them as darkflats and bias but I get the same result sadly. Other flat frames overcorrect all over the place. perhaps I have an issue with my baader MPCC causing a reflection, from my sensor to the MPCC and then back towards my camera? And now that I have changed my system, the new flats don't suffer it so badly?

This is what a single stretched flat looks like for the image at the start of the post, pretty normal I'd say.

[pipnina]

Flat_001.fitsFlat_003.fitsFlat_002.fitsM_81_Light_001.fitsM_81_Light_003.fitsM_81_Light_002.fits

Here's some samples, when I added my darks that matched my lights it didn't change anything so I dunno if it's worth including them.

Bias_002 - Copy.fits Bias_003 - Copy.fits Bias_001 - Copy.fits

[ONIKKINEN]

Yes, you do need darks even if the camera is low noise. Its a pretty clear cut case, examples below:

First calibrated without darks:

Then calibrated with your bias frames as darks (and bias frames):

Now the problem isn't completely gone as i can still see the outline of the 100% illumination zone as the darkened area, because you really would need to use darks to make this work properly.

Another issues to keep in mind would be more mechanical rather than data handling. Did you move the camera before taking the flats? Did the scope sit in sunlight and warm up? Many things can happen with newtonians that ruin flats because the primary mirror can move in the bottom. Thermal differences between night and sunlight could probably also deform the tube enough to take you out of collimation. Flats being offcenter and showing this kind of weird center pattern could be due to that, miscollimation because of either the mirror cell itself being sloppy or the tube deforming, or the focuser being sloppy or really anything. I would try to take flats at night in the middle of the session to make sure they are the best they can be. You can test whether your scope has some mechanical issues by taking one set of flats at the beginning of the session, one in the middle and one in the end. The tube will be in different orientations and so mechanical issues will manifest themselves to the flats. If you see these issues, you need to fix the root problem before flats will ever work consistently.

But you can take darks now and see if it fixes, or makes the issue better (it will). The mechanical thing is somewhat of an eternal migraine for newtonian users, myself included, but think of that later.

 

[ONIKKINEN]

Also, forgot a simple method of testing the flats.

Just look at your calibrated frames from all over the session. All of them should have their flats applied the same way, but if some of them have the donut or some other problem in them and some dont you will know that some subs were in a different collimation (due to anything, focuser, mirror etcetc) than the flats.