Flats Are Overcompensating Vignetting?

@rathijit

Hi,

Did you use bias or dark flat for calibration ? I notice that different bias affects flat field calibration too. 

Kijja

@kijja

I used bias, but just for experimenting, I tried with and without bias frames as well. It resulted in the same weird inverse vignetting issue.

@rathijit

I use qhy 163m which has the same CMOS sensor as your ZWO. The over-corrected flat field in narrow band filters was gone after using 0.3second bias. Shorter biases didn’t work with my camera. 

Kijja

@kijja

Hmm, I will try to generate a new set of biases. However, with this set of biases, I don't face any issues when using a refractor. I thought it might be due to the EdgeHD optics. 🤔 

It is indeed a thing that a slightly longer sub might help with these sensors. The reason why exactly I can't find at the moment, but it's worth a try.

I tried longer (0.3sec) biases, and I even tried Flat Darks with no luck. I keep getting the over-compensated flats.

Another thing I tried was using sky-flats instead of my flat box, but did not have any luck there either.

Adding a link to the data in case someone wants to try this out:

https://www.dropbox.com/sh/sjhhqu0moe6cnza/AACZ4WVRoZdwsYHJxRqrvS87a?dl=0

@rathijit

Just try your data. My result is exactly similar to yours. So, I think flat subs themselves are the cause of vignette over corrected. Still have no idea why flat subs behave like this.

Kijja

By the way, the integrated image without flat calibration looks good, very few amp-glow is seen.

Thanks for the data link! That's always helpful. I think Kijja shows the problem to be with the flat indeed. As to what might be wrong with them exactly can be difficult to figure out from here. It seems the drop in the flat frames on the sides is simply to big compared to when you make the light frames. That would indicate a change in the optical path between the two situations. Is that the case?

@vincent-mod

No, the optical path was identical. I think I am close to figuring out the issue... I retook the bias and dark frames and the flats started working much better. There is still a very light over-compensation, but much better than before, which tells me the flats are not getting calibrated correctly by the bias and dark frames.

That's interesting as well, then the problem would be those frames as all programs you tried apparently calibrate the flats in a similar way. Could you post a masterdark and masterbias from the ones maybe having a problem versus the new ones?

@vincent-mod

Sure, I added it to the previously posted dropbox link, under the "Masters" folder inside "Newly Collected Calibration Frames". The only difference between the old and the new files I discovered was that the camera offset in the old files were set lower. I had collected the old boas and dark frames a long time ago, so I guess I had forgotten what offset value I had set then.

@kijja

 I agree... I have been making do without flats. In fact, the final processed image looks pretty great without flats, but I would love to be able to correct it further with flats. Here is the final integration: https://www.astrobin.com/417848/

I'll get to them later, but you might have figured the problem out already. Maybe playing around a bit more with the offset eliminates it all together..

I liked your image on Astrobin btw, that looks amazing!

@vincent-mod

Yeah, looks like it was the offset, and a lot of folks with the same camera have claimed that using dark flats solves the issue altogether. I'll try all of those.