Flats not correcting dust motes properly...

[Adam1234]

Hi All... could someone please help me identify why my flats aren't correcting properly and advise what I can do to improve the correction? More specifically, I have a series of dust motes, some of which have been corrected, but others are just leaving large ringed artefacts that look sort of like craters on the moon. Here are some stacked images to show you what I mean:

(NGC2403) R filter - calibrated with darks and flats (with dark flat calibration)

(NGC2403) R filter - calibrated with darks only - No flats

 

(NGC2403) G filter - calibrated with darks and flats (with dark flat calibration)

(NGC2403) G filter - calibrated with darks only - No flats

(NGC2403) B filter - calibrated with darks and flats (with dark flat calibration)

(NGC2403) B filter - calibrated with darks only - No flats

(M33) L filter - calibrated with darks and flats (with dark flat calibration)

(M33) L filter - calibrated with darks only - No flats

 

Some notes to help the diagnosis (and hopefully the resolution):

• Equipment: SW Evostar ED80, ZWO ASI1600mm pro, APT for image capture and flats aid, ZWO electronic filter wheel (Set to unidirectional)
• FYI images autostretched & uncropped
• 100 Flat frames for each filter taken with APT flat aid, target ADU 25,000, using the white t-shirt method and white laptop screen
• All dark frames (light darks and flat darks) are matched in exposure length to the respective lights and flats
• Flats were taken the following day after imaging
• All equipment remained on scope in the same position between lights and flat, i.e. did not remove or rotate camera etc
• Calibrated, registered & stacked in Pixinsight
• Darks/Flats/Flat darks taken in the dark to ensure no light leak 

 

Some of my thoughts:

• I thought it could it be where I took all the flats the next day, and therefore the filters weren't in the exact same focus position as the light frames which may cause different sized dust donuts in the lights vs flats. However, I'm not so sure as 
  • 1) I've never had this problem before when taking flats the next day, and have successfully taken flats with all filters in one go before, and it seems most people do this
  • 2) The last filter I used on the night was the L filter, so if this was the case you would expect this to have the best correction, but this is probably the worst affected. In fact I ran ABE on the L image it is absolutely horrible and really shows the artefacts (see image below), but ABE on the other images did not make them worse at all
  • 3) I have an EAF so I couldn't have accidently changed the focus by a considerable amount
• Would trying for longer flat exposure times work? The reason I ask this is because the best correction in my opinion (although maybe not perfect - the rings may possibly still be very very faintly visible if I look closely) is with the B filter, with which the flat frames exposure time was the longest at 7.34375s.  Flat exposure time for L & G = 5.30469s, and R = 5.98438

(M33) L filter - calibrated with darks and flats (with dark flat calibration)_ABE

 

Any advice will be greatly appreciated as being in the South UK I don't get many clear nights as it is, and I would rather not have to bin images I do get on these rare clear skies!

 

Thanks in advance

Adam

 

Edited January 27, 2021 by Adam1234

[Adam1234]

If it helps here are the Master flats L > R > G >B, autostretched

 

Edited January 27, 2021 by Adam1234

[tomato]

My understanding is that the embossed effect on the dust motes is caused by a slight shift in position of the camera between the light and flat frames being taken. Were they taken with a big difference in the ambient temperature, I guess that may be a possibility if you are taking them the next day. I sometimes encountered this effect when I used to set up and take down each time, but not since I have had a permanent setup.

M33 is looking good once you have the calibration problem sorted.

[Adam1234]

4 minutes ago, tomato said:

My understanding is that the embossed effect on the dust motes is caused by a slight shift in position of the camera between the light and flat frames being taken. Were they taken with a big difference in the ambient temperature, I guess that may be a possibility if you are taking them the next day. I sometimes encountered this effect when I used to set up and take down each time, but not since I have had a permanent setup.

That is a good point about ambient temperature - it was sub-zero outside when I took the lights (it is winter after all), and even though I cooled the sensor to the same temperature as the lights, the ambient temperature was ~18C inside when I took the flats. That is obviously quite a difference, but at least is something I can test by re-taking some flats outside (as long as it is not raining!)

I really hope that this is the reason and not that the filter wheel isn't aligning each time the filter is changed as that may be more problematic!

24 minutes ago, tomato said:

M33 is looking good once you have the calibration problem sorted.

Thanks! Fingers crossed I get the problem sorted! I won't get to finish M33 this year as it getting a bit on the low side and too close to streetlights, but there's always next year!

[Adam1234]

As a side note - if I can't sort out the problem, I will probably have to clean my optics - objective lens and filters. I have a load of Zeiss wipes at home, would these be safe enough to use?

https://www.amazon.co.uk/ZEISS-Lens-Wipes-Pack-200/dp/B00IKGH2TI

[Rustang]

9 hours ago, Adam1234 said:

As a side note - if I can't sort out the problem, I will probably have to clean my optics - objective lens and filters. I have a load of Zeiss wipes at home, would these be safe enough to use?

https://www.amazon.co.uk/ZEISS-Lens-Wipes-Pack-200/dp/B00IKGH2TI

I know alot of people swear by the 'Baader Optical Wonder Fluid', Ive got some and its pretty good. Your also best getting some lint free drying cloths aswell to get rid of any excess residue that both your wipes and the wonder fluid may leave behind. Go gentle!

[Rustang]

These are the wipes (drying cloths) I use with the fluid and to remove excess residue .

 

https://www.bristolcameras.co.uk/p-the-dust-patrol-delta-optical-cleaning-wipes-100-pack-.htm

Edited January 29, 2021 by Rustang

[Adam1234]

9 minutes ago, Rustang said:

I know alot of people swear by the 'Baader Optical Wonder Fluid', Ive got some and its pretty good. Your also best getting some lint free drying cloths aswell to get rid of any excess residue that both your wipes and the wonder fluid may leave behind. Go gentle!

Thanks, I'll look into the wonder fluid - I think I've seen it on FLO

[Rustang]

Just now, Adam1234 said:

Thanks, I'll look into the wonder fluid - I think I've seen it on FLO

Its definitely worth having around, and its not that expensive. Ive used it when i needed to clean the front of the lens on my SW ED80 and 72 after a few nights use👍

[Rustang]

Also Ive found that getting the right flats even for a DSLR can be an art form in its self (I'm still not 100% there) I appreciate I don't use the same equipment as you so all I can say is just keeping trying different ways and settings and see what ends up working for you. I would have thought someone with more experience would have been able to help though.

[geeklee]

10 hours ago, Adam1234 said:

As a side note - if I can't sort out the problem, I will probably have to clean my optics - objective lens and filters. I have a load of Zeiss wipes at home, would these be safe enough to use?

https://www.amazon.co.uk/ZEISS-Lens-Wipes-Pack-200/dp/B00IKGH2TI

If you haven't got one already Adam, I can definitely recommend a "rocket blower" (I.e. bulb blower). This will get rid of debris/dust first before going in with any wipes. Alternatively you may find just using this is enough. It's also handy to use inside the filter wheel and flattener etc. 

Agree with @tomato that I've read similar as the reason for the crater type appearance is from the dust moving slightly between lights and flats (although if uses the filter movement you mentioned is also possible and something to keep an eye on)

I try to think any prevention is better than cure especially if you tear up and down like me. Start as clean as you can within reason and where possible keep that image train sealed. 

Edited January 29, 2021 by geeklee

[Adam1234]

3 hours ago, geeklee said:

If you haven't got one already Adam, I can definitely recommend a "rocket blower" (I.e. bulb blower). This will get rid of debris/dust first before going in with any wipes. Alternatively you may find just using this is enough. It's also handy to use inside the filter wheel and flattener etc. 

Agree with @tomato that I've read similar as the reason for the crater type appearance is from the dust moving slightly between lights and flats (although if uses the filter movement you mentioned is also possible and something to keep an eye on)

I try to think any prevention is better than cure especially if you tear up and down like me. Start as clean as you can within reason and where possible keep that image train sealed. 

I have one of those rocket blowers actually - it did nothing to the dust on the front lens, but good shout to try that on the filters first before using any wipes/fluid, as my filters are mounted I can see myself using a wipes and just moving the dust to the sides rather than removing it completely 🙃

Ok, so I think my plan of action at the moment is to retake my flats outside (when it's not raining) so the ambient temperature is similar to that of when I took the lights, and hope that the movement between lights and flat was caused by some contraction/expansion of the optical train caused by the difference in the ambient temperature, and if possible if there is some clear sky (🤣🤣🤣) focus each filter before taking the flats to eliminate any focus issue.

Then if the problem persists, strip down and clean filters.

Fingers crossed that the apparent movement is not caused by the filter wheel not returning to the same position, as I've no idea how to solve that without buying an expensive filter wheel...🤞

[Adam1234]

Ok I took new flats outside last night to see if this had any effect. Also I adjusted the brightness on the laptop screen I used as a light source to increase exposure time to 8.25s for all filters (target ADU 25000). Calibrated the flats with matching darks.

It does not appear that this has made any difference:

 

As another experiment, I calibrated my previous set of flats taken last week with the new set of flats, with my theory being that they should cancel each other out if all is well with the filter wheel alignment, but as shown in the images below there are similar artefacts as in the light frames:

 

 

Does this essentially prove that something is wrong with the filter wheel and that it is not returning to the exact same position each time? 

If this is the case, aside from just cleaning the filters (as inevitably dust will present itself again and I'll have the same issue), is there anything I can do about this? I have had a search around and lots of other people on the ZWO forum have had similar issues with this filter wheel but looking at the responses ZWO seem to be really ignorant to the issue and very unhelpful.

 

Adam

[tomato]

It is looking like a filter wheel position registration issue, especially if this problem has been reported previously with this unit.

I wonder do ZWO use a different technology to register the wheel position compared to other manufacturers of EFWs? Can anything be done to address this or is this inherent in the design? I note in the original post you note that the wheel is set to unidirectional. I assume this means it will rotate in either direction to get to the desired filter quicker, would it be worth running it in one direction only to see if this makes the positioning more reliable (if that is an option).

[Laurin Dave]

If you’ve not done so perhaps  recalibrating the filter wheel will help..  

Dave

[Adam1234]

8 hours ago, tomato said:

It is looking like a filter wheel position registration issue, especially if this problem has been reported previously with this unit.

I wonder do ZWO use a different technology to register the wheel position compared to other manufacturers of EFWs? Can anything be done to address this or is this inherent in the design? I note in the original post you note that the wheel is set to unidirectional. I assume this means it will rotate in either direction to get to the desired filter quicker, would it be worth running it in one direction only to see if this makes the positioning more reliable (if that is an option).

I'm not sure exactly what technology they use to register the position, possibly a laser or some kind of sensor, I'd have to do a bit of digging. Also not sure what other manufacturers use. 

From what I have read online, the unidirectional setting is one direction, and is what I have seen recommended online.

 

[Adam1234]

6 minutes ago, Laurin Dave said:

If you’ve not done so perhaps  recalibrating the filter wheel will help..  

Dave

Good shout, ill give that a go, and take another set of flats to see if that corrects the issue - probably unlikely if the filters were out of position when the lights were taken but it's worth doing.

As another experiment, after recalibrating, I could take one set of flats say in R, change to a different filter, then back to R and take another set of flats, and see if the dust in the 2 sets of flats match, at least that would tell me if the recalibrating worked even if I can't correct my images.

After that I think I will carefully clean the filters ( got some baader wonder fluid on it's way) and hope that i can keep dust out of the filterwheel in future!

[ollypenrice]

One test of the F/W positioning hypothesis would be to calibrate one flat with another from the same run. That would remove the possibility of wheel movement. (Or you could make one master flat from half your set and a second from your second set and see if they can correct each other. Your idea of dividing a flat by a flat is sound. I do this when teaching imaging to show what flats do. If a flat divided by a flat from the same run comes out genuinely flat then positioning will be the culprit.

Sometimes flats problems can be utterly baffling.

Olly

Edited February 3, 2021 by ollypenrice

[ollypenrice]

2 minutes ago, ollypenrice said:

One test of the F/W positioning hypothesis would be to calibrate one flat with another from the same run. That would remove the possibility of wheel movement. (Or you could make one master flat from half your set and a second from the other half of your set and see if they can correct each other. Your idea of dividing a flat by a flat is sound. I do this when teaching imaging to show what flats do. If a flat divided by a flat from the same run comes out genuinely flat then positioning will be the culprit.

Sometimes flats problems can be utterly baffling.

Olly

 

[Adam1234]

3 hours ago, ollypenrice said:

One test of the F/W positioning hypothesis would be to calibrate one flat with another from the same run. That would remove the possibility of wheel movement. (Or you could make one master flat from half your set and a second from your second set and see if they can correct each other. Your idea of dividing a flat by a flat is sound. I do this when teaching imaging to show what flats do. If a flat divided by a flat from the same run comes out genuinely flat then positioning will be the culprit.

Sometimes flats problems can be utterly baffling.

Olly

Hi Olly, thank you for your input, that will be a great test. I took 100 flats for each filter so plenty to split in half and create 2 masters from. I'll try that this evening and report back!

👍

[Adam1234]

Well, it looks to me like the filter positioning is certainly the culprit.

As per @ollypenrice's suggestion, I made a master flat from one half of a set of flats and another master from the other half taken in the same run (50 flats in each half), and calibrated one with the other. This is my result with the L flats (filter with the most dust):

 

Looks pretty flat to me with none of the 'moon crater' artefacts I'm seeing with the lights. The R, G & B flats all look corrected as well and look exactly like above so no point posting those.  

So I think that this confirms that my flat frames themselves are fine (reassuring in some respects), and that the issue is due to the filter wheel not returning to the same position with enough accuracy.

Ok, now I need to figure out why it's not returning to the same position, and how I can fix this. 

 

Maybe I ought to try this:

1) Re-calibrate the filter wheel (I believe there is an option in the ascom driver)

2) Take a set of flats > change to a different filter > change back to the original filter > take another set of flats

3) Stack each set of flats and calibrate one master with the other

4) Jump with joy when the flats correct each other indicating problem has been solved; or

5) Start crying, chuck the filter wheel in the river and buy a new one (not ZWO) / seek further advice from the pros

 

Adam

 

[Laurin Dave]

I’d just do 1-4 if I were you..  😉 and as another thought what filters are you using (some are too tall and rub on the faceplate) and are they securely located 

[Adam1234]

32 minutes ago, Laurin Dave said:

I’d just do 1-4 if I were you..  😉 and as another thought what filters are you using (some are too tall and rub on the faceplate) and are they securely located 

I'll try and avoid step 5 if at all possible 😀 

I'm using the ZWO 1.25" LRGB Sii Ha Oiii filters that came as a bundle with the filter wheel and ASI1600, so shouldn't be an issue with the filters 

 

 

 

[ollypenrice]

What about sending the FW back? Or asking the supplier for their opinion?  In the scheme of things, getting a wheel to stop in about the right place should not be diificult.

Olly

[Adam1234]

34 minutes ago, ollypenrice said:

What about sending the FW back? Or asking the supplier for their opinion?  In the scheme of things, getting a wheel to stop in about the right place should not be diificult.

Olly

Good idea, I'll try recalibrating first and if that doesn't work then I'll contact FLO who I bought it from and see if they can help