Flats not working. All out of ideas

[geordie85]

Since this lovely clear spell has ended I finally ran some new flats, but the problem is they're over correcting and I've tried everything I can think of.

I took them exactly the same way as I have always done but this time they're not working, so I have alot of data I can't use.

I use a zwo 183mm, astrodon LRGB gen 2 filters, Esprit 100 with matching flattener and a dimmable tracing pad and white paper.

So far I've tried ADU 30000 taken at 0.5 seconds, 1.17 seconds and 2.5 seconds with a 5 second delay between subs.

ADU 22000 0.5 and 1.4 seconds again with a 5 second delay and all are temperature matched to my lights.

I've tried using different amounts between 30 and 100. All with matching dark flats and the correct darks for my lights. Tried stacking in astroart 7 and in DSS. Also tried average and sigma in astroart.

I set my scope up and left it out in the same place the entire time I was imaging and covered up during the daytime.

I'm stumped and don't want to throw away so much data.

Any more ideas anyone?

[Stuart1971]

Hmmm, it seems that you are doing it all correctly, the only thing I can add, is that CMOS cameras seem to work better with slightly longer flats, mine are all 4 seconds, and 25,000 ADU…

[Same old newbie alert]

Can you post a eg of your flats?

 

[geordie85]

25 minutes ago, Stuart1971 said:

Hmmm, it seems that you are doing it all correctly, the only thing I can add, is that CMOS cameras seem to work better with slightly longer flats, mine are all 4 seconds, and 25,000 ADU…

Exactly, that's why I'm stumped. I've never had a problem before, even with short exposures.

1 minute ago, Same old newbie alert said:

Can you post a eg of your flats?

 

I will do in a couple of hours when I'm home

[Same old newbie alert]

Are you taking flats to get rid of vignette, or dust bunnies

 With a 183 most use dark flats🌛

[Stuart1971]

6 minutes ago, Same old newbie alert said:

Are you taking flats to get rid of vignette, or dust bunnies

 With a 183 most use dark flats🌛

You use dark flats to calibrate the light flats, Dark flats on there own, are just darks, or bias….you need light flats with every camera AFAIK

Edited March 28, 2022 by Stuart1971

[rickwayne]

I  use bias, not flat darks, with my 183. And with my LED tracing panel I wind up with sub-second exposures even at gain 53. Maybe they're too short but eh, it's good enough for me (check my Astrobin if you wonder if that's good enough for you. Not bragging -- just calibrating).

I agree that it doesn't sound as if there's something wrong with your method. I'd be suspicious of the processing.  Some notions I'd have to run down the problem:

• What exactly is the problem you're seeing when you process with these flats?
• Is the problem something you can examine by stretching a flat sub-exposure, or the integrated master flat?
• Do you have any of your old flats still lying around?  Do they look similar when stretched? Other than the dust bunnies being all wrong, what happens if you swap them into  your processing?

[geordie85]

38 minutes ago, rickwayne said:

I  use bias, not flat darks, with my 183. And with my LED tracing panel I wind up with sub-second exposures even at gain 53. Maybe they're too short but eh, it's good enough for me (check my Astrobin if you wonder if that's good enough for you. Not bragging -- just calibrating).

I agree that it doesn't sound as if there's something wrong with your method. I'd be suspicious of the processing.  Some notions I'd have to run down the problem:

• What exactly is the problem you're seeing when you process with these flats?
• Is the problem something you can examine by stretching a flat sub-exposure, or the integrated master flat?
• Do you have any of your old flats still lying around?  Do they look similar when stretched? Other than the dust bunnies being all wrong, what happens if you swap them into  your processing?

My main problem is there is 1 massive dust bunny that always over corrects and the background is horrible. I've not got round to processing as i don't see the point if my flats aren't working correctly. After stacking Astroart shows an auto stretched image and its always there.

I've also just stacked in Astro Pixel Processor and still getting the same thing. 

Here is an auto stretch from Astro Pixel Processor so you can see what i mean (i'm not bothered about the edge difference as i would crop beore any proper processing). Saved at JPEG at 100% quality for here. Also attached is a single flat frame

[david_taurus83]

The big one looks like it's in an inverted position on the calibrated light stack. If it's bottom right on the flat, it also looks like it's top left on the stack. I would inspect all your lights and see if something has been flipped around or inverted for whatever reason.

[geordie85]

2 hours ago, david_taurus83 said:

The big one looks like it's in an inverted position on the calibrated light stack. If it's bottom right on the flat, it also looks like it's top left on the stack. I would inspect all your lights and see if something has been flipped around or inverted for whatever reason.

The stacked image is 196 lights taken over 2 nights. It's the same even if I just stack 1 nights worth of data.

[Shibby]

The larger circles don't really look like dust bunnies to me - they might be something related to the filters... The flats don't look quite right. Is it possible that light is leaking in somewhere? Were the flats taken in daytime? Do you have the imaging train covered up?

[AstroNebulee]

4 hours ago, david_taurus83 said:

The big one looks like it's in an inverted position on the calibrated light stack. If it's bottom right on the flat, it also looks like it's top left on the stack. I would inspect all your lights and see if something has been flipped around or inverted for whatever reason.

I too thought I could see the offending dust bunny smudge inverted in the light frames 

[david_taurus83]

2 hours ago, geordie85 said:

The stacked image is 196 lights taken over 2 nights. It's the same even if I just stack 1 nights worth of data.

Have you got Pixinsight? You could Blink through all the light frames and see if the circles stay in the same place. 

[geordie85]

1 hour ago, Shibby said:

The larger circles don't really look like dust bunnies to me - they might be something related to the filters... The flats don't look quite right. Is it possible that light is leaking in somewhere? Were the flats taken in daytime? Do you have the imaging train covered up?

Flats were taken both at night and during the day.  If memory serves right the above flat was taken at night, but I'm not 100% sure. I'll upload one I know I took at night when I get home. What sort of filter issue could cause this?

3 minutes ago, david_taurus83 said:

Have you got Pixinsight? You could Blink through all the light frames and see if the circles stay in the same place. 

Unfortunately no I don't. I have flicked through them all in DSS though and it appears to be in the same place in all my subs. 

[Shibby]

10 minutes ago, geordie85 said:

What sort of filter issue could cause this?

I really don't know - it's always incredibly difficult to figure these things out. But there could be some sort of reflected, stray light, that at some point has passed (possibly backwards) through a filter causing the dark circle. That stray light could easily come from the light panel if it was at night.

[geordie85]

10 minutes ago, Shibby said:

I really don't know - it's always incredibly difficult to figure these things out. But there could be some sort of reflected, stray light, that at some point has passed (possibly backwards) through a filter causing the dark circle. That stray light could easily come from the light panel if it was at night.

That it is. I'm usually pretty good at figuring these problems out but this time I'm beat. Hence asking the collective here.

Strange thing is on my last night I took flats at the beginning and at the end of my imaging session and the data collected that night has the same issue. So it cannot be a moving dust bunny

[rickwayne]

If I look reeealy closely I can see a much fainter big circle on the flat corresponding to the lower-right problem. So perhaps not an orientation issue after all.

You could run the numbers on the diameters of the circles to see how far the offending thing is from the sensor. I mean, if you find a chunk on the corresponding surface you've only fixed a symptom, but knowing where it's coming from could be instructive. 

[geordie85]

9 hours ago, rickwayne said:

If I look reeealy closely I can see a much fainter big circle on the flat corresponding to the lower-right problem. So perhaps not an orientation issue after all.

You could run the numbers on the diameters of the circles to see how far the offending thing is from the sensor. I mean, if you find a chunk on the corresponding surface you've only fixed a symptom, but knowing where it's coming from could be instructive. 

Thanks.

It seems that no matter what, my data cannot be calibrated with the flats I'm taking so I'm going to cut my losses and disassemble the entire rig, give it a good cleaning and hope for some more clear nights again soon.

Thanks all for the advice and suggestions.

[rickwayne]

Sorry man.

[geordie85]

14 minutes ago, rickwayne said:

Sorry man.

Don't be. You didn't do it haha.

To be fair before I started I swapped my camera from my samyang to my scope. I should have given it a clean before hand.

I've now taken the whole thing apart, cleaned the sensor window, filters, flattener and lens. I'll run some new flats when I've got the time and hopefully there'll be no moving dust bunny. I assume it was moving since flats wouldn't take care of it.

It's just a shame I've lost so much data.

Once again thanks for all the advice.

Edited April 1, 2022 by geordie85