Taking effective Lum flats with a cheap newtonian

[Uranium235]

The principle aim of this idea is to reduce the amount of unwanted, scattered light when taking flats with a cheap newt, the Explorer range of newts to be exact - the same could apply to any other newt that doesnt have baffles and just has the bog standard synta spray job. 

While all is spiffing in narrowband, the 130pds had a little problem with producing a decent Lum image - with the problem being unable to get a suitable flat frame with a large CCD camera. The method of getting the flats hadnt changed since my 80ED, which was to have the panel or laptop screen flush to the OTA with a few sheets of A4 to dim and diffuse it. However, for the 130pds - this seemed to cause issues with stray light completely bypassing the primary/secondary and going straight up the drawtube, reflecting off the filter, back on to the secondary, then back to the CCD. 

Having read the SBIG paper on flat fields the other night, it got me thinking about how to best replicate a sky flat - as that has been the only method that has produced a good Lum flat so far. The paper referred to "dome flats" as a good alternative, but since I dont have an obsy it never occured to me to use a large monitor backed away from the telescope a few feet (as it would be inside an obsy dome)

The solution was to use my 22" LED monitor as a lightbox, then use the opaque plastic from my broken GN flatfield panel to cover the 130pds apeture (hung off it like a b-mask). The monitor was about 5ft away from the scope, and Al's virtual lightbox was set at about 80% brightness. Additional measures to cut out stray light include binbagging the bottom of the scope, doing it in a very dark place (the cellar), and making sure there are no other sources of light in there (including power LEDs).

So, here is a comparison of the two methods: (Using a Lum filter as NB is not affected)

First, the light source flush against the OTA:

As you can see, there is a clear circular reflection/artefact in the flat frame - quite ugly.

Next, the with a light source backed off and diffused with the GN spare part:

Looks much better!

Now if we apply those flat frames to 20min worth of the cocoon, we can see the kind of chaos it was causing. Both images have had a single stretch in levels, just to show up any background issues.

Bad flat applied:

Good flat applied:

Hopefully, this observation will help others who have encountered the same problem when taking flats. The nub of it being, with a cheap newt - dont have the light source flush against the OTA!

And here is the same image again, except a couple more subs and processed a bit:

[ollypenrice]

A very informative post on a difficult subject.

Olly

[Uranium235]

Thanks Olly, its some of your words that were echoing about in my head when I put this together (regarding interior paint and the spectra of LED panels, blue heavy)...lol. 

Initially, I thought it was all down to how the spacing was configured, or the corrector itself - but Im relieved its now sorted. And what it does mean is that I can now chase a bit of dark stuff without worrying about dragging up background issues when chasing the weaker signals.

[tingting44]

Excellent write up, will try this tonight doing some test flats for the first time and see what results I get

[Atreta]

Can you post how it would look like with no flats applied?

great image.

[Thalestris24]

Hi Uranium

Thanks for posting this.

I use my widescreen monitor with a Notepad window for taking flats. I position it about a meter or so away from the scope. Flats have seemed ok but I've only used them for osc images. I might do mono images at some point so will bear this in mind Not that I'm ever likely to get such quality images - you must have much darker skies than me. Mind you, I think most people have darker skies than me!

Cheers

Louise

[Uranium235]

Can you post how it would look like with no flats applied?

great image.

I'll have a rummage around for the raw subs in a bit (theyre on the PC in the basement), but without calibration theyre pretty much unuseable.

[brendant]

An excellent observation and we'll demonstrated with pictures, I am sure this will help many imagers improve their skills in the dark art of flat fielding, thank you for sharing.

Cracking image of the Cocoon.

B.

Sent from my Nexus 4 using Tapatalk

[Uranium235]

Ahhh here we go, here is a stack of the subs - completely raw, no calibration, just a crude stretch in levels:

[Thalestris24]

Hiya

So that's mostly/all just vignetting?

Louise

[Uranium235]

Hiya

So that's mostly/all just vignetting?

Louise

Yep, thats about it. It only really starts to drop off towards the corners. But that would probably have something to do with the size of the secondary mirror as no part of my optical train could be considered restrictive (its 2" all the way to the camera).

[Thalestris24]

Yep, thats about it. It only really starts to drop off towards the corners. But that would probably have something to do with the size of the secondary mirror as no part of my optical train could be considered restrictive (its 2" all the way to the camera).

Hiya

Same here - I thought I'd mostly eliminate it by using all 2"/M48 connectors but it's still there. Sigh. Maybe also something to do with having a big chip? I hadn't thought about the secondary... Mind you, I also haven't flocked my scope - not sure if internal light reflections might contribute?

Louise

[Uranium235]

Hiya

Same here - I thought I'd mostly eliminate it by using all 2"/M48 connectors but it's still there. Sigh. Maybe also something to do with having a big chip? I hadn't thought about the secondary... Mind you, I also haven't flocked my scope - not sure if internal light reflections might contribute?

Louise

Flocking does not contribute to a reduction in vignetting. But taking flats completely cures the problem, and its part of the normal process of image capture. Even if you had a larger secondary, you would still have to flat field.

[ollypenrice]

Yes, you cannot image properly without flats. The most powerful of the software fixes is DBE in Pixinsight but flats are a fact of imaging life.

Olly

[Thalestris24]

Of course!

[edarter]

Hi, just seen this topic linked from another,  very interesting as I've only just started imaging and while I've done darks and bias I've not generated any flats yet.

I do have a question though, my monitor isn't the largest, so would a smart TV with a plain white background on the web browser do the same thing? It's a much bigger surface area than my monitor

Thanks

Ed