Please advise regarding C8 Flats not calibrating dust motes

oymdRostokko
55 replies316 views
oymd avatar

Hi everyone

Yesterday night I imaged M63.

2600MC Pro, Celestron C8 with 0.63x reducer, OAG-L and 174MM Mini

All using NINA. Flats taken with my regular Geoptik Manual Flat Panel.

Lights: 85 subs, 300s, Gain 100, Offset 20, Temp -5C

I took my set of Flats & Dark Flats at about 4am, at the end of the Light images run. Nothing was touched on my image train.

Flats: Avg ADU: 29,000, Gain 100, Offset 20, Temp -5C. Flats were 1.69s.

I ran the data through WBPP, and immediately saw that my Flats did not calibrate the dust motes or the vignetting out.

Attached is a screen shot showing my FLATS MASTER and LIGHTS MASTER.

Only GRAXPERT has been applied to the LIGHTS MASTER to highlight the failed calibration.

Any advice as to why my Flats are not calibrating correctly?

Many thanks

I was not able to attach the XISF Master Light and Master Flat? How can I share them to the forum?

Screenshot 2026-03-31 135414.png

Helpful Respectful Engaging
nico1038 avatar

Hi @oymd

The embossed donuts are very carcateristics of flats that are offset from the lights.

Did you use a filter for this session? If so, how was it integrated into the optical train?

Nico

oymd avatar

IDAS LPS D2 filter.

It’s one of 5 filters I have in a ZWO 5×2” EFW

nico1038 avatar

oymd · Mar 31, 2026, 12:09 PM

IDAS LPS D2 filter.

It’s one of 5 filters I have in a ZWO 5×2” EFW

So you have a filter wheel. Has it rotated between the lights and the flats?

Well Written
oymd avatar

nico1038 · Mar 31, 2026 at 12:12 PM

oymd · Mar 31, 2026, 12:09 PM

IDAS LPS D2 filter.

It’s one of 5 filters I have in a ZWO 5×2” EFW

So you have a filter wheel. Has it rotated between the lights and the flats?

No. Image train was untouched. When NINA completed the LIGHTS sequence, I opened Flat Wizard, which slewed to the Zenith, put the Geoptik flat panel on the scope, and ran the flats wizard.

Helpful
nico1038 avatar

oymd · Mar 31, 2026, 12:14 PM

nico1038 · Mar 31, 2026 at 12:12 PM

oymd · Mar 31, 2026, 12:09 PM

IDAS LPS D2 filter.

It’s one of 5 filters I have in a ZWO 5×2” EFW

So you have a filter wheel. Has it rotated between the lights and the flats?

No. Image train was untouched. When NINA completed the LIGHTS sequence, I opened Flat Wizard, which slewed to the Zenith, put the Geoptik flat panel on the scope, and ran the flats wizard.

Are you sure about that? Because that's often the cause of this kind of problem (the wheel, after rotating, doesn't stop in exactly the same place).

In any case, this means that there is play somewhere in the optical train (probably in the wheel itself, even if it is not rotating) and that the filter has moved slightly while the telescope was pointed towards the zenith.

Well Written
oymd avatar

nico1038 · Mar 31, 2026, 12:22 PM

oymd · Mar 31, 2026, 12:14 PM

nico1038 · Mar 31, 2026 at 12:12 PM

oymd · Mar 31, 2026, 12:09 PM

IDAS LPS D2 filter.

It’s one of 5 filters I have in a ZWO 5×2” EFW

So you have a filter wheel. Has it rotated between the lights and the flats?

No. Image train was untouched. When NINA completed the LIGHTS sequence, I opened Flat Wizard, which slewed to the Zenith, put the Geoptik flat panel on the scope, and ran the flats wizard.

Are you sure about that? Because that's often the cause of this kind of problem (the wheel, after rotating, doesn't stop in exactly the same place).

In any case, this means that there is play somewhere in the optical train (probably in the wheel itself, even if it is not rotating) and that the filter has moved slightly while the telescope was pointed towards the zenith.

Thanks for your input Nico

I agree, it must have to do with the carousel in the EFW.

Oh, wait, you are onto something.

When NINA ends a sequence, I use the legacy sequencer, it parks the scope at the end of the imaging run, and the EFW rotates the carousel back to position 1, which in my case is NO FILTER.

So, when I ran FLAT WIZARD, I unparked first, then chose LPD-D2, which is in position 2.

So, the carousel definitely moved!

How can I fix this then? In the settings, I have already chosen UNIDIRECTIONAL movement, but I never clicked the CALIBRATE toggle since I bought it and added it to the image train?

Should I try the CALIBRATE thing? And what des it actually do?

Lastly, if our guess is right, shouldn’t the dust motes show up on the flats master BUT AT A DIFFERENT ANGLE OF ROTATION?

Many thanks

Ossi

Respectful Engaging
Miguel A. avatar

Are the filters mounted (screw in) or unmounted? I have the mounted filters and I have found when I first installed them into the EFW they were a little lose, resulting in a shift. If this applies to you, take a look at the filters and tighten them in the filter holder.

Helpful Concise
Rostokko avatar

Are you 100% positive those flats are shot using that filter? That master looks suspicious to me.

You can easily verify that by opening one of the flat exposures and checking the filter recorded by nina in the fit headers.

nico1038 avatar

oymd · Mar 31, 2026, 12:28 PM

How can I fix this then? In the settings, I have already chosen UNIDIRECTIONAL movement, but I never clicked the CALIBRATE toggle since I bought it and added it to the image train?

The unidirectional option would have been my next suggestion, so I'm not really sure how to help you anymore. To be honest, I don't have much experience with filter wheels. Others will undoubtedly have better advice.

oymd · Mar 31, 2026, 12:28 PM

Lastly, if our guess is right, shouldn’t the dust motes show up on the flats master BUT AT A DIFFERENT ANGLE OF ROTATION?

The difference in angle is probably very small but it creates a sufficient misalignment of the spots to produce this effect (with donuts having one light side and the other dark, giving this 3D effect).

If you can see the dust spots on your lights, you should be able to notice the offset by blinking betwen your masterflat and a light frame.

GalacticRAVE avatar

My suspicion would be that you have a bit of slag/backlash in your image train. Basis for this reasoning: you have large dust mots (ie dust on the filter), small dust mots (dust on the camera window) and variations in vignetting. If it would be the filter wheel (if it does not find exactly the same filter position - common problem in particular for bidirectional EFWs), you would only have amboss imprints for the big mots but not for the sensor (no change whatever the filter position is) or the vignetting. CS Matthias

Helpful
nico1038 avatar

GalacticRAVE · Mar 31, 2026, 12:45 PM

My suspicion would be that you have a bit of slag/backlash in your image train. Basis for this reasoning: you have large dust mots (ie dust on the filter), small dust mots (dust on the camera window) and variations in vignetting. If it would be the filter wheel (if it does not find exactly the same filter position - common problem in particular for bidirectional EFWs), you would only have amboss imprints for the big mots but not for the sensor (no change whatever the filter position is) or the vignetting. CS Matthias

That's an excellent point. I hadn't thought of that!

oymd avatar

Thank you all for your suggestions.

I am using 2” mounted filters.

After a week of bad weather, yesterday night was my first session of imaging for a while.

I assembled the train yesterday. The 2600mc pro is bolted to the EFW. The OAG is bolted on the other side of the EFW.

The only connection that MIGHT be loose is the M48 adapter between the OAG and the SCT adapter screwed on to the 0.63x reducer.

I just checked it. It’s rock solid.

Again, I’m a little confused.

If the theory that the carousel was at a slightly different position when doing the flats as opposed to the lights, wouldn’t the flat master show the motes but at a different position?

Lastly, the flats are registered in NINA as being LPS-D2, as expected.

Many thanks again.

Well Written Respectful
Rostokko avatar

Was the camera temperature at the time you shot flats equivalent to the one while shooting lights? Are you using flat darks? If yes, is wbpp associating the correct darks to the flats in question? And while in wbpp, can you triple check that the lights are indeed associated to the intended flats?

Well Written Helpful Concise Engaging Supportive
oymd avatar

Rostokko · Mar 31, 2026, 01:49 PM

Was the camera temperature at the time you shot flats equivalent to the one while shooting lights? Are you using flat darks? If yes, is wbpp associating the correct darks to the flats in question? And while in wbpp, can you triple check that the lights are indeed associated to the intended flats?

Yes to all of the above.

I always check in WBPP CALIBRATION tab, and everything was mapped correctly.

I have copied the details off my first post above:

Lights: 85 subs, 300s, Gain 100, Offset 20, Temp -5C

I took my set of Flats & Dark Flats at about 4am, at the end of the Light images run. Nothing was touched on my image train.

Flats: Avg ADU: 29,000, Gain 100, Offset 20, Temp -5C. Flats were 1.69s.

Dark Flats: Gain 100, Offset 20, Temp -5C. Dark Flats were 1.69

I ran the data through WBPP, and immediately saw that my Flats did not calibrate the dust motes or the vignetting out.

There are no other FLATS or DARK flats in the folder to confuse WBPP.

Helpful
oymd avatar

I have just dismantled the image train, opened the EFW, and opened up NINA.

The EFW works perfectly, and all filters when chosen in NINA center perfectly onto the sensor.

Well Written Respectful Concise
Rostokko avatar

oymd · Mar 31, 2026, 02:13 PM

Flats: Avg ADU: 29,000, Gain 100, Offset 20, Temp -5C. Flats were 1.69s.

Dark Flats: Gain 100, Offset 20, Temp -5C. Dark Flats were 1.69

I know I am being annoying - but I’m also sure that’s why you posted the original question here: you are hitting a wall based on what you know, and looking for some other angle to investigate. Been there, done that.

So, I’ll be annoying one more time: did you shoot dark flats in the morning? Or have they been shot separately, in a controlled environment; I suppose the real question is: are dark flats “really” dark?

Well Written Respectful Engaging Supportive
nico1038 avatar

Just a quick demonstration that it is not a calibration problem, but rather a mechanical one.

You can recreate the embossed donuts effect by dividing a flat by itself but with a shift of 20 pixels in both direction.

exemple.jpg

These kinds of donuts are so distinctive that they leave little room for doubt, in my opinion: something has moved between the acquisition of the lights and the flats. The question is to understand what .

Well Written Helpful Insightful Respectful Concise Engaging
oymd avatar

Rostokko · Mar 31, 2026, 02:30 PM

oymd · Mar 31, 2026, 02:13 PM

Flats: Avg ADU: 29,000, Gain 100, Offset 20, Temp -5C. Flats were 1.69s.

Dark Flats: Gain 100, Offset 20, Temp -5C. Dark Flats were 1.69

I know I am being annoying - but I’m also sure that’s why you posted the original question here: you are hitting a wall based on what you know, and looking for some other angle to investigate. Been there, done that.

So, I’ll be annoying one more time: did you shoot dark flats in the morning? Or have they been shot separately, in a controlled environment; I suppose the real question is: are dark flats “really” dark?

You’re not being annoying at all?!

Why did you get that impression?

I’m grateful for the help and advice.

Yes, come to think of it, the Flat Wizard was run at dawn, so there was light out there, but I ran the flat Wizard UNDER the Telegizmos 365 cover, and had a look at the Dark Flats with ASI Studio, and the histogram was all the way to the left, like what I am accustomed to see with Dark Flats.

I’ll attach some pics of the FLAT & DARK FLAT subs.

I wish there is a way to upload the actual XISF MASTERS?

The question I keep asking, but no one is answering: if the issue was that something moved between imaging and Flats, why is my Flat Master so clean and not showing ANY dust motes at a slightly different position?

oymd avatar

nico1038 · Mar 31, 2026, 02:31 PM

Just a quick demonstration that it is not a calibration problem, but rather a mechanical one.

You can recreate the embossed donuts effect by dividing a flat by itself but with a shift of 20 pixels in every direction.

exemple.jpg

These kinds of donuts are so distinctive that they leave little room for doubt, in my opinion: something has moved between the acquisition of the lights and the flats. The question is to understand what .

Thanks for that.

Whilst I do not understand the science behind your statement, I’m happy to go with your theory.

I’m not really experienced, and think you might have got to the root of the issue.

Do you think it might be the mirror in the SCT the shifted ever so slightly when I slewed to the zenith in Flats Wizard in NINA? Could that be the cause of the very slight movement between the light images and the Flats?

also, why doe the Flats Master not show those two large dust motes in the center, but slightly in a different place if the movement theory is the one to go with?

Thanks again!

Respectful Supportive
Rostokko avatar

oymd · Mar 31, 2026, 02:38 PM

The question I keep asking, but no one is answering: if the issue was that something moved between imaging and Flats, why is my Flat Master so clean and not showing ANY dust motes at a slightly different position?

Visually comparing images (auto)stretched at different levels can be very misleading - I think that’s why nobody is answering that question explicitly.

@nico1038 is making a very good case, I believe, for mechanical shift; what perplexes me about that is that it is a fairly big shift, and I have never seen anything like that with my C8 - which does mess around with the efw to adjust focus every hour or so before it gets to the point of shooting flats. That said, the point made there is very true.

If you want to upload xisf masters (why not light too?), you probably want to use a cloud drive of your choice and include links here.

Engaging
nico1038 avatar

oymd · Mar 31, 2026, 02:43 PM

also, why doe the Flats Master not show those two large dust motes in the center, but slightly in a different place if the movement theory is the one to go with?

The files would be really helpful: couldn’t you use a cloud service and share a link here?

Alternatively, can you show us your masterflat again but after debayering it and applying an unlinked automatic stretch?

Well Written Helpful Respectful Concise Engaging Supportive
oymd avatar

Rostokko · Mar 31, 2026, 02:48 PM

oymd · Mar 31, 2026, 02:38 PM

The question I keep asking, but no one is answering: if the issue was that something moved between imaging and Flats, why is my Flat Master so clean and not showing ANY dust motes at a slightly different position?

Visually comparing images (auto)stretched at different levels can be very misleading - I think that’s why nobody is answering that question explicitly.

@nico1038 is making a very good case, I believe, for mechanical shift; what perplexes me about that is that it is a fairly big shift, and I have never seen anything like that with my C8 - which does mess around with the efw to adjust focus every hour or so before it gets to the point of shooting darks. That said, the point made there is very true.

If you want to upload xisf masters (why not light too?), you probably want to use a cloud drive of your choice and included links here.

Fab…I have Dropbox…Not sure how to do it, but I will attach link to:

Master Light, with a couple of Light subs

Master Flat, with a couple of FLAT subs and DARK FLAT subs…

I suppose the MASTER DARK is not relevant here?

nico1038 avatar

oymd · Mar 31, 2026, 02:51 PM

I suppose the MASTER DARK is not relevant here?

I don't think it's a calibration issue, but if I'm wrong, then the MasterDark is relevant. Add it if you can.

oymd avatar

Rostokko · Mar 31, 2026, 02:48 PM

oymd · Mar 31, 2026, 02:38 PM

The question I keep asking, but no one is answering: if the issue was that something moved between imaging and Flats, why is my Flat Master so clean and not showing ANY dust motes at a slightly different position?

Visually comparing images (auto)stretched at different levels can be very misleading - I think that’s why nobody is answering that question explicitly.

@nico1038 is making a very good case, I believe, for mechanical shift; what perplexes me about that is that it is a fairly big shift, and I have never seen anything like that with my C8 - which does mess around with the efw to adjust focus every hour or so before it gets to the point of shooting flats. That said, the point made there is very true.

If you want to upload xisf masters (why not light too?), you probably want to use a cloud drive of your choice and include links here.

https://www.dropbox.com/scl/fi/oy5tniqb4hpg7gkipbpo8/Investigating-C8-Flats.rar?rlkey=4lv9ryx6ggs5xitqmlxi52hod&st=6krls86q&dl=0

I have created a WinRAR archive containing 4 folders:

  • Masters Folder:

Master Dark 1.69s

Master Flat 1.69s

Master Light 300s

Master Light 300s autocrop_1x drizzle

  • Light Folder:

2 Light subs 300s each

  • Flats folder:

2 Flats subs out of 50

  • Dark Flats folder:

2 Dark Flats subs out of 50

Please advise if I have done this correctly and you have access to the files.

Many thanks!!