Flats causing headaches

[Lee_P]

So, here's the thing: when imaging with my DSLR I could get decent flat files no problem. But since moving to an ASI1600 I haven't cracked it. I can get flats that cancel out dust donuts, but the vignetting is still there. I've tried flats with means between 3500 and 20,000. Maybe it's an issue with the settings I'm using in DSS..?

This is the best result I can get after stacking with flats:

Any pointers would be appreciated!

 - Lee

[sloz1664]

Hi Lee,

What are you using to obtain your flats?

Steve

[Lee_P]

T-shirt flats with a cloudy sky; against a laptop screen on white; and against a big TV displaying white. Similar results all round. I can never crack the vignetting.

[Gerry Casa Christiana]

It could be as you say settings in DSS. I would reset settings to default and run it again. It doesn't seem feasible to be anything else. 

Gerry

[carastro]

Can you upload or make a link say in dropbox to the files = lights, flats and bias, so we can have a go and see whether its the stacking software, as that will help to eliminate that.

Carole 

[Lee_P]

4 hours ago, carastro said:

Can you upload or make a link say in dropbox to the files = lights, flats and bias, so we can have a go and see whether its the stacking software, as that will help to eliminate that.

Carole 

Thanks for the offer, here are some of the files: https://www.dropbox.com/sh/kdmc02hplmxwcg1/AACAgEni5mmocImx2tDDKthFa?dl=0

I've tried different settings in DSS, but no joy. So I'm either missing something, or my flat files are fundamentally flawed somehow.

[carastro]

Well that was fun.  Astroart threw a hissy fit with those subs and I eventually found out that half of them must have been after a meridian flip so had to rotate them 180 degrees.

The ADU on the individual flats were 5712 - 5800.  I always do them around 24,000/25,000 or as near as I can get.  

Yet the stacked flat once stretched appears to be OK.  So am puzzled why.  Give it a try with flats at the correct ADU and see what happens.

Don't think it can be the stacking software as that's two different softwares give the same result. 

I am more concerned with the fact that the flats seem to have what I call a tyre effect.  i.e. the corners are darker and then there is a lighter circle with a slightly less darker middle, and I am wondering whether it is your method of capture that is causing the problem, perhaps the illumination is not even, try just using the sky and a T shirt.   I noticed evidence of this on your original post. 

I've attached the stacked and processed result, and the stretched master flat, see what others make of it.

Carole 

 

[carastro]

Oh, and by the way, if you look closely the flats did not remove the dust on my processed image, but that could be because the flats are upside down as I didn't know which way up to do them.

Carole 

[Stub Mandrel]

1 hour ago, carastro said:

Well that was fun.  Astroart threw a hissy fit with those subs and I eventually found out that half of them must have been after a meridian flip so had to rotate them 180 degrees.

If half of the images are being rotated, they will need the flats to be rotated as well... as the vignetting is not perfectly central, I guess that may be the problem.

[Tim]

12 minutes ago, Stub Mandrel said:

If half of the images are being rotated, they will need the flats to be rotated as well... as the vignetting is not perfectly central, I guess that may be the problem.

Could be, yes. Carole, if you are up for it, try stacking just the pre-flip images and the flats.

[carastro]

Yes, you are right, stupid me, I should have known that, but still doesn't solve the question of not removing the vignetting.

Will have to download again as I can't remember which ones were flipped. 

Carole 

[martin_h]

If you are using DSS for stacking it will automatically rotate flipped sub's so your problem might not lie there.

[carastro]

Have separated the pre and post subs and processed each separately.  

Still bad vignetting on both.  (reminder I used Astroart for stacking). 

I still think the ADU might be at least part of the problem.  

Carole 

 

 

[carastro]

One other thought, you did not supply any bias subs, I think it would be a good idea to add them too as I gather they can help the flats to be applied properly.

Carole 

[newbie alert]

21 hours ago, Lee_P said:

T-shirt flats with a cloudy sky; against a laptop screen on white; and against a big TV displaying white. Similar results all round. I can never crack the vignetting.

I'm sure there be too much contrast in a cloudy sky, clear blue evening or morning sky is best..

[Stub Mandrel]

I use an indirectly-lit by even natural light matt painted wall about a metre from the end of the scope. The short distance ensures no surface marks can be remotely in focus, and I rotate the scope through a full circle as I take the flats.

[carastro]

Well since every-one is offering their methods, I used to use an EL panel until it went wrong, while I was waiting for it to be repaired I waited until the following day and did sky flats, but I always put a piece of typing paper over the aperture +/- a white pillowcase folded and strapped flat and just fiddle around with the ADU until I get the right reading.

I point at the sky in the opposite direction from the Sun and since I use the typing paper and pillowcase to diffuse the light it seems to work just fine. 

In the event, my EL panel was irreparable, and I decided to stay with Sky flats.  

Carole 

 

[Stub Mandrel]

I just put my camera on Av setting and it works brilliantly - I know dedicated astro cameras don't have simple auto exposure...

[carastro]

Quote

I just put my camera on Av setting and it works brilliantly - I know dedicated astro cameras don't have simple auto exposure...

I used to use AV and 100 iso on my old Canon 450D and as you say it worked brilliantly.  However that died on me after several years use, and so I replaced it with a Canon 1100D - imagine my horror when that method did not work on the 1100D, had to devise a new method completely by taking an exposure in Bulb of something like 1/80 sec, but this varies every time depending on the brightness of the sky, whereas AV on the 450D just used to adjust accordingly. 

I also use CCD cameras and the same applies I have to get the right ADU (for each filter as well).  Gone are the easy days of 450D and AV flats.

Carole 

[Lee_P]

On 11/13/2017 at 09:38, carastro said:

One other thought, you did not supply any bias subs, I think it would be a good idea to add them too as I gather they can help the flats to be applied properly.

Carole 

Sure, I've added a folder with some bias: https://www.dropbox.com/sh/kdmc02hplmxwcg1/AACAgEni5mmocImx2tDDKthFa?dl=0

Thanks very much for helping with this!

[carastro]

I don't know why but I can't seem to download the bias to dropbox EXPLORE, they appear in Dropbox but not in the Explore folder.  

Carole 

[Lee_P]

30 minutes ago, carastro said:

I don't know why but I can't seem to download the bias to dropbox EXPLORE, they appear in Dropbox but not in the Explore folder.  

Carole 

I've uploaded them to another Dropbox folder, in case that helps.

https://www.dropbox.com/sh/a55cqdrkzf9b77j/AAAJEPFPzYRaDGTcq0wsHa4ma?dl=0

[ollypenrice]

It might be instructive to take one single flat sub and divide it by another different flat sub. The result ought to be a flat (but noisy) frame. 

Olly

[MartinFransson]

How do I fint out the ADU of a sub? I'm about to take flats with my 1600MM as well. 

[Lee_P]

On 11/18/2017 at 09:10, MartinFransson said:

How do I fint out the ADU of a sub? I'm about to take flats with my 1600MM as well. 

If you're using SGP Pro, you can load an image and then click on Image Statistics.