Artefacts left in master that flats are not clearing up..... Help please

[stevek]

Hi there, could really use some suggestions please folks?

I posted here my master luminance flat and resulting master luminance on M101 - 15 x 10minute subs.  Can't seem to see how I embed them into the body of this message on the forum??

I have a couple of strange arcs in the image - look to the left of M101 at the 9-10  o'clock position on this picture and see the long, faint arc.  This seems to correspond to one of the fainter but larger bunnies on the flat but it is not being removed.....  I have this issue with other pictures too and it is ruining my astrophotography and I need to sort it out. Was working fine so not quite sure what's changed.....  Maybe I will try manual pre-processing to see if that sorts it out .  I am using BPP.

Flats taken at *approximate* (not identical because how can you do that with autofocus?)  focus position of the lights and no movement at all in the rig and focuser.  Camera is same temp.
1.  Take flats first (whilst getting darker - pretty dark when I take them with obs roof on)
2.  Start lights
3.  Meridian flip occurs and lights capture continues.

In Pixinsght I add all sub frames and I select one - the best - light to act as the reference.

It's my understanding that only one set of flats required and that registration will sort it all out i I don't need different flays either side of the Meridian???  Yet, I get some strange arcs and artefacts still.  Should the reference frame correspond to the orientation of most of the lights or does it not matter?

Any ideas please?
Steve (at wits end)

[stevek]

Any thoughts please?
Here is another, side by side showing the flat and stacked master and the artefact that's left.  There are others too if you look closely.

[stevek]

Maybe my flats are too bright and the flat is being over subtracted (or added)   I have the identical problem with another image set - M51 this time.
The raw subs off of the camera do not show the large faint arcs, only the smaller bunnies and these smaller bunnies are correctly being removed by calibration.  However it appears the larger (and fainter) arcs are being *introduced* by image calibration into the calibrated lights.  So it is at the calibration stage this problem is occurring, not image integration.

This suggests to me that there is something wrong with my flats - they are too bight or too faint - they are 25000 ADU average.   Fortunately, I have not moved the camera on the scope so can retake a set with lower and another set of higher ADU to see if that solves the problem.

I am on latest version of Pixinsight and the same problem occurs on either of my 3 PC's or my 2 MAC's.

[pfile]

if that arc is the leftover from that large dust spot, more than likely the dust mote moved between the flat and light acquisition. if the dust motes are large, they can travel around.

www.wilmslowastro.com/software/formulae.htm

that site has some formulas that will help you figure out where exactly the dust motes are (distance from sensor) based on your OTA. you can then judge how likely they are to move around (for instance, if they are on the filters, there's a good chance they migrate as the filter wheel turns, or maybe your filter wheel is not perfectly centering the filters each time around).

brightness of the flat should not matter that much, unless you're looking at a situation where you have reflections in the tube that manifest only at higher brightness.

rob

[stevek]

Thanks for your input.

I tried doing flats with different ADU from 7000 all the way up to 30000.  Same result.  I then used a different illumination source and used a panel directly at the end of my scope tube as opposed to the EF panel on the wall (which I just built).  Same again, I can't get rid of the large arc dust bunnies.   Then again, this has now happened on two sets of images -I also have this on an M51 too.  It can't be the same on two sets of images on two different nights.....  Ill have to bin 20 hours of data -precious in the UK with our constant rain....

Only workaround I can think of is to use a blur filter to see if I can process it out.

Only thing that is new on the scope - and I have never had this issue before with this scope and camera combination - is a new FF I installed on my TEC140 two months ago since we have had endless bad weather in UK and not had time to use it since.  But that's the whole point of the flat fields - to eliminate these optical effects.

That said, I did these flats a few days after the lights for various reasons (but not moved anything on the scope).  And, thinking about it, I tut-tutted myself the other day for not covering the objective lens up for two days......  Hmmmm, food for thought.  Could dust have got on the objective - and being a long way from the sensor (and hence really large bunnies) the flats done two days later - from an uncovered objective -  do not now match the lights and thus introducing this issue? I wonder....  I will try rocket blowing the objective and will try again.  But there is a lesson here - make sure to do the flats immediatly or at least the following morning....

[pfile]

if you use those formulas youll see that a piece of dust on the objective casts a shadow larger than the sensor area...

one other possibility, if these are not simply shifted dust motes, is that you have reflections inside the OTA, possibly off of the FF/FR. i know i have problems with that on both of my telescopes. you can try sky flats to see if that changes anything...

anyway the only thing i can suggest is to take a light and then immediately take a flat and see if it flattens properly. if so, then the problem is the motion of dust motes.

rob

[stevek]

After closer inspection I see that the "arc bunnies" are on the individual lights as well - faint but visible - as per the composite image below.  So it can't be a moving dust mote, Pixinsight can't be calibrating the lights properly or the flats are wrong (but how?!!! I tried various lightness settings).  Calibration is taking out the small bunnies but not the large scale structures on the larger ones.  All I am using is the BPP with default settings.  I may try doing this manually like I used to do.

[pfile]

why cant it be a moving dust mote? the flats just dont match, that happens all the time with moving motes. its not likely to move during light or flat acquisition, it happens inbetween as the telescope is slewing or the filter wheel is moving.

rob

[blueridgeDSIA]

Have you tried creating an roi over one of the areas and fine tuning the structure rejection settings in image integration?