After months of trying I cannot calibrate my RGB images!

I performed a full manual calibration workflow, and the resulting master light seems pretty good to me. The master flat has some star streaks which I wasn't able to reject due to the low frame count, so there are some corresponding dark streaks in the calibrated lights, but the dust filaments and motes calibrated out perfectly.

Untitled 2.png


In detail what I have done:
  1. Integrated the flat darks into a master flat dark (no normalization, no weighing)
  2. Integrated the darks into a master dark (no normalization, no weighing)
  3. Calibrated the flats with the master flat dark
  4. Integrated the calibrated flats into a master flat (multiplicative normalization, no weighing, equalize fluxes)
  5. Calibrated the lights with the master dark and the master flat
  6. Registered the calibrated lights
  7. Integrated the registered calibrated lights
 
If I look closely, I maybe see one faint dust mote that moves from one light frame to the next, at 10 o'clock from the galaxy. Just maybe.
 
Yes, but the minute you integrate 50 of these, they show up like crazy. Thanks.
 

Attachments

  • 50 subs.JPG
    50 subs.JPG
    392.9 KB · Views: 32
As @fmeschia said it is very hard to see anything with the images you provided. Can you upload the 50?
If i had to guess I'd say that the dust spots moved at some point during your acquisitions.
 
yeah 100% - that image in post #24 is the hallmark of a dust mote that's moved, or a filter that's not repeatably returning to the same position. if you can't get rid of the spot the best thing to do is take the camera off the telescope and clean the filters (based on the size i assume the dust is on the filter, but that's something you'll have to calculate.
 
What's puzzling is that I see at least three "layers" of dust, so at least one of them is not on the filters. But all three layers show the tell-tale signs of moving dust.
 
There is no way that the dust has moved due to the filter wheel or any couplings in use. The filter wheel has been changed 3 times, a ZWO, a ATIK and now a high end FLI. Always the same. Very rigid couplings (68 mm).
There are dust motes on the cover glass (inside that appeared after changing desiccant). Those show up as well (to the right of center about 1/2 way down from the galaxy.). Thus that has nothing to do with the filter wheel.
So why are all the narrowband working but not RGB? If it was movement, it should apply there as well.
All that can be left is the focuser ( a moonlight chl) but how could that affect anything.
 
tomb18 said:
What does that mean? It's just 5 instead of 50 lights.
Click to expand...
It means that if the problem shows only after a large number of lights are integrated, providing only 5 lights doesn't help forming much of an opinion or diagnosis.
 
tomb18 said:
There is no way that the dust has moved due to the filter wheel or any couplings in use. The filter wheel has been changed 3 times, a ZWO, a ATIK and now a high end FLI. Always the same. Very rigid couplings (68 mm).
There are dust motes on the cover glass (inside that appeared after changing desiccant). Those show up as well (to the right of center about 1/2 way down from the galaxy.). Thus that has nothing to do with the filter wheel.
So why are all the narrowband working but not RGB? If it was movement, it should apply there as well.
All that can be left is the focuser ( a moonlight chl) but how could that affect anything.
Click to expand...
The big, dominant dust donuts here are, by my estimate, 900 pixels wide. With your setup that places the occluding objects 24mm from the sensor.

Are you using an external focuser and a mirror locking system? Because with an SCT, unless the mirror is solidly locked, it can rock enough to shift the light path, which is optically similar to the position of the dust motes changing. Which is what we see visually as that embossed appearance, light one side and dark on the other.
 
cloudbait said:
The big, dominant dust donuts here are, by my estimate, 900 pixels wide. With your setup that places the occluding objects 24mm from the sensor.

Are you using an external focuser and a mirror locking system? Because with an SCT, unless the mirror is solidly locked, it can rock enough to shift the light path, which is optically similar to the position of the dust motes changing. Which is what we see visually as that embossed appearance, light one side and dark on the other.
Click to expand...
I have uploaded all 51 subs. Big file (5Gb). https://drive.google.com/file/d/1itbwrppTSqatcuQM3tVKwqrMin5GZAPs/view?usp=sharing

I am using a Moonlight CHL external focuser and I tighten the mirror as much as I can. Still this does not explain why I never have issues with narrowband. Furthermore, if the mirror is shifting I should see a different pattern during focusing. The poisson dot is perfectly centered all the time that I focus no matter the position in the sky.
The 24mm distance corresponds to the actual filters. I chose the green one here, since it has the most dust motes. The other long defect is on the camera window.
 
tomb18 said:
I have uploaded all 51 subs. Big file (5Gb). https://drive.google.com/file/d/1itbwrppTSqatcuQM3tVKwqrMin5GZAPs/view?usp=sharing

I am using a Moonlight CHL external focuser and I tighten the mirror as much as I can. Still this does not explain why I never have issues with narrowband. Furthermore, if the mirror is shifting I should see a different pattern during focusing. The poisson dot is perfectly centered all the time that I focus no matter the position in the sky.
The 24mm distance corresponds to the actual filters. I chose the green one here, since it has the most dust motes. The other long defect is on the camera window.
Click to expand...
I think the mirror could easily move enough to cause what we're seeing an not show up in the focus pattern.

In any case, I'm not clear at this point what flats and darks to tie to these lights. Are they available?
 
cloudbait said:
I think the mirror could easily move enough to cause what we're seeing an not show up in the focus pattern.

In any case, I'm not clear at this point what flats and darks to tie to these lights. Are they available?
Click to expand...
I have uploaded the flats and darks here: Disregard that the darks say RED in the name of the files. That's just what Voyager has named the darks.

Images - Google Drive

drive.google.com drive.google.com

I could upload all the flats and darks but I might run out of space.
 
tomb18 said:
Haven't tried that. Raining today so I will do this next. What would be the outcome?
Click to expand...

i am wondering if the illumination profile you're getting from the panel flat is different from the sky flat (which could in turn be different from the lights themselves...)
 
tomb18 said:
I have uploaded the flats and darks here: Disregard that the darks say RED in the name of the files. That's just what Voyager has named the darks.

Images - Google Drive

drive.google.com drive.google.com

I could upload all the flats and darks but I might run out of space.
Click to expand...
At the least your flat is badly underexposed, which could be at least part of the problem.
 
No it was much worse at 32000 ADU. This got rid of some of the reflections going to 12000 ADU
 
You must log in or register to reply here.
Back
Top