Taking Flats with a DSLR

[MarkS]

Or think about it this way:

Draw a squiggle in black felt pen on the objective lens of your refractor.  Illuminate the objective lens.  The lens with its squiggle becomes an object for your focal reducer to create an image.  Find out the focal length of your reducer and then use the the lens equation to find out where the image of this squiggle will appear.

Then puzzle about it for a while and await the subsequent flash of inspiration.

Mark

[Ian]

Right, I couldn't find a black marker so I used silicon carbide scribe.

I'm now sat patiently awaiting a flash of brilliance. Just so I know, what's one like?

As another mind exercise I've tried to remember the last time you were wrong about something. The best I could come up with was when you said "here, try some of this weird French drink".

What I'm now thinking about is the requirement to have something to establish is an image (squiggle, stripes etc) and you argument then becomes a discussion on the effects of faults in the lightbox design.

This therefore brings me back to where I was before. Is there a difference between a focussed image of a perfectly uniform diffuse light source that is of effectly infinite dimensions and one that is out of focussed.

Thinking harder time again...

[PhilB]

Mark, your argument appears to boil down to: a light-box flat produces an out of focus image that is much further back behind the plane of the infinity focus than does a does a sky flat. In this case, distortion increases with distance, hence a light-box flat is more distorted than sky flat. Am I getting there?

[Rocket Pooch]

Lets not also talk about wavelengths or this will never end.

[MarkS]

This therefore brings me back to where I was before. Is there a difference between a focussed image of a perfectly uniform diffuse light source that is of effectly infinite dimensions and one that is out of focussed.

Yes, there is a difference.  Although your diffuse light source may be infinite, your imaging train has finite sized lenses, mirrors, baffles etc.
It is the finite sized lenses, mirrors, baffles etc. that cause the vignetting.  The vignetting (which may have multiple causes) is what you are trying to measure with your flat.  

As you move your infinite uniform diffuse light source nearer or further from the scope it does alter the raypaths to the CCD and the distribution of light on that CCD.

Mark

[MarkS]

Mark, your argument appears to boil down to: a light-box flat produces an out of focus image that is much further back behind the plane of the infinity focus than does a does a sky flat. In this case, distortion increases with distance, hence a light-box flat is more distorted than sky flat. Am I getting there?

Yes - I do think you have just about grasped it!

Mark

[MarkS]

Lets not also talk about wavelengths or this will never end.

I don't this will ever end anyway !

[PhilB]

Mark, your argument appears to boil down to: a light-box flat produces an out of focus image that is much further back behind the plane of the infinity focus than does a does a sky flat. In this case, distortion increases with distance, hence a light-box flat is more distorted than sky flat. Am I getting there?

Yes - I do think you have just about grasped it!

OK, so the reason that the problem presents itself more aggressively when a focal reducer is present is due to the reducer stretching the focal plane and, thereby, stretching any distortions that may be present too. Hence, the light-box flat suffers more by this process because it's distortions are greater in the first place.

[Rocket Pooch]

Oh no!

[PhilB]

Brace yourself, Chris - I haven't got to my question about wavefronts yet 

[Rocket Pooch]

Hang on let me add to that.

What you are suggesting is that there is an imperfection caused before and after the focal reducer is added?

If it is present after and is more aggressive then the issue is irrelevant, a flat will still cure this if it is taken correctly with the components all in line.  

Example a) I have some dust, it looks like this

O O O
O O
O

This is without the focal reducer, and say it has an effect of 5% deviation over against the light frames.


Example b) with the focal reducer it looks like this (smaller o's)

ooo
oo
o

I have reduced the image by 50% to get better S/N but the dust is not a 10% deviation over the light frames.  But they have better signal, maybe more noise depending on your location.

If I take accurate flats for example a) or b) can fix both issues.

I still think the issue Mark has shown in the HH image might be because the systems were setup different, or the light box was poorly built.


Oh and just for fun light pollution

As in my setup on one of my images is say the flat has a deviance of 1% of the overall image and the image has good signal to noise ratio then the dust is mostly insignificant.

In a light polluted area and the defects are more significant in your original images due to high background noise adding to the dust effect.

See example below;

No flat


NoFlat by chrissuddell, on Flickr

Flat


O3Flat by chrissuddell, on Flickr

Flat frame


Flat by chrissuddell, on Flickr

[Rocket Pooch]

Brace yourself, Chris - I haven't got to my question about wavefronts yet 

Oh thats easy to answer.

[MarkS]

I'm now beginning to think I might be wrong about this after all.

Whilst it's true that the focal reducer can form an image of the objective lens (or of the primary mirror), this fact might be a complete red herring 

[Rocket Pooch]

But, the most interesting thing in all this is it would be quite a good thing to do is write a paper on light pulluted subs and flats and non-light polluted subs and flats.  The number of and why use them is always discussed.

Chris

[Rocket Pooch]

Oh and Robert, before you say my image is out of focus it was Mark's fault for bringing that French S*&t to DSC.