Lightbox for Flats

[carastro]

I am surprised that Vlaiv says 3/4 of the way across for the histogram as I always have been told 1/3 across and that has always worked for me.  When I do my flats for the CCD cameras I use an ADU of around 22,000 which is 1/3 of the full well depth of 65,000, but Mo you can;t measure it that way with CR files.

I looked back at some old DSLR flats I did and they were just under halfway.  So might be an idea to try that as a starting point, but I definitely agree with Vlaiv that they are under exposed.

Unfortunately flats are probably the hardest thing to get right in Astrophotography for beginners. 

Old DSLR flats screen shot in Digital Photo Professional (part of the Canon Utility suite.  

Edited December 23, 2019 by carastro

[RolandKol]

Ok, so I took some time and stacked them all.

I have autoremoved background from Light without any cropping and autostreached them using delinger script on Pi...

Single subs did look nice and even master Flat looks even, but after correction it kills the right side of the images...

AutoStreached Dark

 

Flat

 

Single Light

 

Light (stacked)

 

And the last, very interesting one, is Higher pixerl rejection map, which is genereted by Pi image integration tool

 

[RolandKol]

Will stack without flats soon and see what we have.

 

[RolandKol]

Found the issue...
Decided to Blink all lights before stacking again...

Problem hides in the Lights.... Cloud came in....

half of them should be deleted before stacking....

Blink is generating the video at the moment, - will upload it soon also

Edited December 23, 2019 by RolandKol

[vlaiv]

7 minutes ago, RolandKol said:

Problem hides in the Lights.... Cloud came in....

half of them should be deleted before stacking....

Are we talking about same set of subs posted earlier?

How do you explain my calibration with single sub and the fact that central part is over corrected while edges are under corrected?

[Marmo720]

12 minutes ago, RolandKol said:

Ok, so I took some time and stacked them all.

I have autoremoved background from Light without any cropping and autostreached them using delinger script on Pi...

Single subs did look nice and even master Flat looks even, but after correction it kills the right side of the images...

AutoStreached Dark

 

Flat

 

Single Light

 

Light (stacked)

 

 

 

 

Thanks @RolandKol. Interesting that the master flat looks fine but the stacked image has one side looking so bad. I would think the issue might have been due to lightbox having LED on one side but why doesn't that show in the master flat? Is it possible the issue is with my lights instead of the flats? 

I think I will do the flats that @vlaiv recommended with both the lightbox and the iPad methods just to rule that out as a potential issue. 

@carastro - a little bit frustrating to stay up all night imaging and to find out that it can't be used due to flats issue. But I am learning loads and looking forward to solving this and hopefully producing some good images. The last good images I took were the Witches Broom and M27 and both had stubborn dust bunnies which we finally resolved. So I am sure this will be too and excited for when it finally is all working  

[RolandKol]

2 minutes ago, vlaiv said:

Are we talking about same set of subs posted earlier?

How do you explain my calibration with single sub and the fact that central part is over corrected while edges are under corrected?

I cannot explain that, - have no relevant knowledge...

Will stack half of the session later  on and see that we have. PI now generating the Bling movie and I am cooking....  a bit too much of multitasking

[Marmo720]

Just now, RolandKol said:

I cannot explain that, - have no relevant knowledge...

Will stack half of the session later  on and see that we have. PI now generating the Bling movie and I am cooking....  a bit too much of multitasking

That is interesting and I really hope it was just a cloud and I can salvage/learn from this. But I think I also had the same issue with M42 which was much later in the evening. I will check once I get home from work in a few hours. 

@vlaiv If almost 50% have a cloud then it is possible you chose single sub with a cloud? Also, this might be a combination issue where single subs show the issue with my under exposed flats while full stacked image shows the impact of the clouds more? 

[vlaiv]

55 minutes ago, Marmo720 said:

That is interesting and I really hope it was just a cloud and I can salvage/learn from this. But I think I also had the same issue with M42 which was much later in the evening. I will check once I get home from work in a few hours. 

@vlaiv If almost 50% have a cloud then it is possible you chose single sub with a cloud? Also, this might be a combination issue where single subs show the issue with my under exposed flats while full stacked image shows the impact of the clouds more? 

I don't think that clouds have anything to do with it. Clouds can make target dimmer and background brighter - but all of that is light reaching sensor and subject to any vignetting / dust shadows. Flat calibration should work regardless if one is shooting clouds or clear skies and there should be no vignetting and dust shadows visible in the calibrated sub. Perhaps uneven background but not vignetted.

[Marmo720]

16 minutes ago, vlaiv said:

I don't think that clouds have anything to do with it. Clouds can make target dimmer and background brighter - but all of that is light reaching sensor and subject to any vignetting / dust shadows. Flat calibration should work regardless if one is shooting clouds or clear skies and there should be no vignetting and dust shadows visible in the calibrated sub. Perhaps uneven background but not vignetted.

That is fair point and would explain why I could see the same issue on different target several hours later. 

Interested to see if Roland had luck with some of the subs and why that might be.

[RolandKol]

Blink vid here

https://youtu.be/Z03XYj-6h6c

The more I look at it... the more I think it was frost..... Not sure....

50% stack incomming

P.S. it is my first Blink vid, - not sure how to increase the quality...

on PI it looks a bit better

Edited December 23, 2019 by RolandKol

[vlaiv]

37 minutes ago, RolandKol said:

Blink vid here

https://youtu.be/Z03XYj-6h6c

The more I look at it... the more I think it was frost..... Not sure....

50% stack incomming

P.S. it is my first Blink vid, - not sure how to increase the quality...

on PI it looks a bit better

I agree, there is some strange effect - but I don't see it be on front of the lens - it will not spread like that what ever shape it has on front of the lens - things will only get darker and that is about it.

It needs to be somewhere inside system - flattener probably as sensor is not cooled and it is warmer than rest of the gear and it will not allow for frost / dew to form.

[RolandKol]

and 50% stack is here

as we can see, we had 2 problems at once.

Lights subs were taken during clouds or the frost was in place

plus, - Flats not doing their job as underexposed....

Maybe something else, - but I would start from here.

 

[RolandKol]

The one in the front is the stack of all subs

Edited December 23, 2019 by RolandKol

[RolandKol]

17 minutes ago, vlaiv said:

I agree, there is some strange effect - but I don't see it be on front of the lens - it will not spread like that what ever shape it has on front of the lens - things will only get darker and that is about it.

It needs to be somewhere inside system - flattener probably as sensor is not cooled and it is warmer than rest of the gear and it will not allow for frost / dew to form.

yep... 

it may be on the flattener as the sensor was around 17C all the time...
strange...

Edited December 23, 2019 by RolandKol

[RolandKol]

just in case if @Marmo720 you want to look at them directlly,

I have placed them here

https://we.tl/t-NGasNtAx9s

[carastro]

It's possible that more than one thing is going on, but in the first instance I think we have all agreed the flats are not right.

Let's get them right first and then see what we are left with.  The horribly stretched image with gradients is what some of my images look like in luminance from home (Bortle 8 ) when the flats are Fine and this is just horrible LP.  Mo lives not so far from me so LP is also an issue here.

I've invited Mo to come over after Christmas with his scope, laptop and camera and I'll sit beside him and see what is happening. 

Carole 

[Marmo720]

 

I had a quick go in between rain showers to take some flats with the lightbox and iPad using white papers. I aimed for about 75% on the histogram and used ISO800. Each flat was taken 5 times so you will see 5x5 flat attempts. 

Hopefully this is useful in diagnosing a potential solution. Clear ish skies tonight so will try with lights too 

https://we.tl/t-0s7e1vQpx3?src=dnl

[RolandKol]

1 hour ago, Marmo720 said:

 

I had a quick go in between rain showers to take some flats with the lightbox and iPad using white papers. I aimed for about 75% on the histogram and used ISO800. Each flat was taken 5 times so you will see 5x5 flat attempts. 

Hopefully this is useful in diagnosing a potential solution. Clear ish skies tonight so will try with lights too 

https://we.tl/t-0s7e1vQpx3?src=dnl

result is even more insteresting...
have you moved your camera since the last sesion?

P.S. stacked iPad+LightBox with 1 sheet only.

Flat still looks underexposed...
Next time use Full ligh on the lightbox and 2 or Max 3 sheets.
if it was already on full power, stay with one sheet and increase exposure a bit.

Light box RGB histogram channels were concentrated in one  peak nicely, - iPad was spread. I would stick to the lightbox.

[Marmo720]

5 minutes ago, RolandKol said:

result is even more insteresting...
have you moved your camera since the last sesion?

P.S. stacked iPad+LightBox with 1 sheet only.

Flat still looks underexposed...
Next time use Full ligh on the lightbox and 2 or Max 3 sheets.
if it was already on full power, stay with one sheet and increase exposure a bit.

Light box RGB histogram channels were concentrated in one  peak nicely, - iPad was spread. I would stick to the lightbox.

Thanks @RolandKol. Yes the camera was originally removed for storage and attached it now for testing and tonight. Probably out of focus too. 

I had both ipad and lightbox at max brightness so interesting there is under exposure. Can you see any gradient for the light box due to led on one side or is it even? 

[vlaiv]

Could you upload a few bias subs with ISO 800?

I did not keep old ones and old upload was deleted by WeTransfer (it expired)

[Marmo720]

26 minutes ago, vlaiv said:

Could you upload a few bias subs with ISO 800?

I did not keep old ones and old upload was deleted by WeTransfer (it expired)

Will do later tonight when I get home. 

Thanks,

Mo

[carastro]

I'm just stacking the image again with the new flats you have made and sent me.  Doubtless since you don't have a permanent set up the camera has been removed so the dust won;t line up, but if the vignetting works we know we are on the right track.

You sent me 3 different versions, so will do a different stack with each.

Carole

[Marmo720]

Is it still under exposed? Even with the iPad and lightbox at maximum brightness and with just 1 white paper, I could only get the combined histogram at about 50%. I will try to do the flats manually and aim for 75% peak. 

At least if the new histogram are less under exposed then we are heading in right direction...

[RolandKol]

53 minutes ago, Marmo720 said:

Is it still under exposed? Even with the iPad and lightbox at maximum brightness and with just 1 white paper, I could only get the combined histogram at about 50%. I will try to do the flats manually and aim for 75% peak. 

At least if the new histogram are less under exposed then we are heading in right direction...

if histogram was around 50%, that will do the job, - but simply make more of them... not 5 but 50.