Hello - I have an issue with my EFW where the luminance (and as far as I know only the luminance) filter placed in slot 1 does not calibrate out correctly between flats and lights. I recalibrate the filter wheel at the beginning of each session, make sure unidirectional rotation is used, and take flats immediately afterward and it is clear the flats are shifted with respect to the lights. I'm wondering if the return position of the wheel is affected by the orientation of the scope since flats are taken with the scope pointing vertically up, whereas lights could have other orientations. The easy solution is to buy another filter wheel, but I mainly wanted to confirm this is an isolated issue with my device and others do not have such calibration problems.
Odd problem with my filter wheel
Well Written Insightful Respectful Concise Engaging
What telescope do you use? You got any images how the flats fail calibrate? Is you filterwheel fully equiped or does a filter or two missing in you wheeel?
The Scope is a 200mm f/4 Newtonian currently with an ASI533MM. All 8 slots are in use. Here is the image showing the calibrated light and the flat. Even at this scale you can see how the artifacts are offset.
Arun H:
The Scope is a 200mm f/4 Newtonian currently with an ASI533MM. All 8 slots are in use. Here is the image showing the calibrated light and the flat. Even at this scale you can see how the artifacts are offset.
Thanks. This problem is only with L? Not with RGB? In what bortle do you image? Let me guess: B6-7? I got the same problem with my 12 in f4. It was a massive primary mirror shift. The mirror was moving in the cell, and even more in the endring of the tube. You can check if your mirror is moving too, by simply push against it from behind the tube or check for any movement.
For some reason flats needs to be perfect in higher bortle, because the vignetting is very strong on the single frames. I dont think you filterwheel is causing the problem. You can check that too, by simply only image L in one night and not rotate your filterwheel positions. Then take flats.
Helpful
Marc Fischer:
Thanks. This problem is only with L? Not with RGB? In what bortle do you image? Let me guess: B6-7? I got the same problem with my 12 in f4. It was a massive primary mirror shift. The mirror was moving in the cell, and even more in the endring of the tube. You can check if your mirror is moving too, by simply push against it from behind the tube or check for any movement.
For some reason flats needs to be perfect in higher bortle, because the vignetting is very strong on the single frames. I dont think you filterwheel is causing the problem. You can check that too, by simply only image L in one night and not rotate your filterwheel positions. Then take flats.
This is both patently untrue and not at all relevant with the case of the OP. If it were the mirrors how on Earth is the dust donut shifting? Certainly isn't attached to either mirrors, is it? There are only two cases worth considering in my view, one is that actually something is in the way either in the stepper motor or the confirmation led (if it has one) which assumes the dust is on the filter, the other is that the dust more is on the sensor and it is this that its shifting for whatever reason. One simple way to check is to compare lights taken with the RGB filter and the one with the L filter and see whether the dust mot shadow coincides or else.
Thank you, Marc, this is very helpful. I do image in B6-7. Thinking about it, yes, a mirror shift would absolutely explain this, since the orientation of the light rays would be correspondingly offset with respect to the filter wheel. You are also correct that luminance in LP is most sensitive to flats, that has been the case with every scope I have used. So then, it isn't that the other filters don't have the offset, it is just that the calibration is less sensitive.
So, as a first step, I will retake the flats in the same orientation as the lights in the hope of salvaging some of this data, but the longer term fix is to carefully tighten the mirror. It gives me a good excuse as well to clean the mirror, something I have been putting off.
So, as a first step, I will retake the flats in the same orientation as the lights in the hope of salvaging some of this data, but the longer term fix is to carefully tighten the mirror. It gives me a good excuse as well to clean the mirror, something I have been putting off.
Well Written Insightful Respectful
andrea tasselli:
This is both patently untrue and not at all relevant with the case of the OP. If it were the mirrors how on Earth is the dust donut shifting?
Hi Andrea, the way I am thinking of this a loose mirror (or tilted mirror) would shift the angle of the converging light with respect to the filter wheel and hence cause the shadow of the dust donut to fall on a slightly different spot on the sensor. Certainly it is possible that the other points you raised are true as well.
Arun H:
Hi Andrea, the way I am thinking of this a loose mirror (or tilted mirror) would shift the angle of the converging light with respect to the filter wheel and hence cause the shadow of the dust donut to fall on a slightly different spot on the sensor. Certainly it is possible that the other points you raised are true as well.
Hi Arun,
I had similar problems and it wasn't down to a shifting mirror but to the position of the filter with respect to the sensor (but mine was a manual FW, so it is understandable). Just shift the flat by the amount required to perfectly overlap the dust shadow on the light and see whether flat-fielding will null the shadow. If you are accurate enough it either does or you'll have a residual. Even if the cone of light from the primary slightly shift of few hundreds of mm the effect is zero on the effectiveness of the flat.
Supportive
andrea tasselli:
Just shift the flat by the amount required to perfectly overlap the dust shadow on the light and see whether flat-fielding will null the shadow.
I did actually do that using a PixInsight script, and yes, it does do a better job at correcting. It creates a new problem though, since the PRNU is shifted between the flat and the light!
Arun H:andrea tasselli:
Just shift the flat by the amount required to perfectly overlap the dust shadow on the light and see whether flat-fielding will null the shadow.
I did actually do that using a PixInsight script, and yes, it does do a better job at correcting. It creates a new problem though, since the PRNU is shifted between the flat and the light!
Still better than the alternative, is it not? And the law of averages will take care of some of it. But, if anything, I think it shows it is the filter position that is the issue.
andrea tasselli:
Still better than the alternative, is it not? And the law of averages will take care of some of it. But, if anything, I think it shows it is the filter position that is the issue.
I think to solve this in systematic fashion, I will do the following:
- Retake the flats in the same orientation as the lights. If I do not do this, there is zero hope of recovering this data (absent accepting PRNU).
- I will check the looseness of the primary mirror in the cell as well as the holder itself. I suspect the holder itself is fine, the mirror in the cell may be a different issue.
- I will open the filter wheel, check for any looseness, and run it and see if it is measurably repeatable.
- I will retighten the primary after cleaning it (since it needs cleaning after two years anyway)
- If the above all do not solve the problem, I will replace the filter wheel.
If I do solve it, I will update this thread so hopefully, the learning is useful to others.
@Marc Fischer @andrea tasselli - thank you both for the sugggestions!
Helpful Supportive
Arun H:andrea tasselli:
Still better than the alternative, is it not? And the law of averages will take care of some of it. But, if anything, I think it shows it is the filter position that is the issue.
I think to solve this in systematic fashion, I will do the following:
- Retake the flats in the same orientation as the lights. If I do not do this, there is zero hope of recovering this data (absent accepting PRNU).
- I will check the looseness of the primary mirror in the cell as well as the holder itself. I suspect the holder itself is fine, the mirror in the cell may be a different issue.
- I will open the filter wheel, check for any looseness, and run it and see if it is measurably repeatable.
- I will retighten the primary after cleaning it (since it needs cleaning after two years anyway)
- If the above all do not solve the problem, I will replace the filter wheel.
If I do solve it, I will update this thread so hopefully, the learning is useful to others.
Hello Arun,
wish you the best! Good luck! You can check if the focuser is shifting too, depending it what direction you look (before and after the flip) Problems like we have are only very noticeable in our bortle :/ B4 and better users never run into issues with flats. I am so jealous
Marc Fischer:
Problems like we have are only very noticeable in our bortle :/ B4 and better users never run into issues with flats
Marc - there are MANY reasons to be jealous of those with B4 and better skies. Lack of problems with flats is just one of them
Dark bands at top and bottom of Minicam8M
EDIT: QHY support responded with the following statement: “We can basically confirm that this phenomenon stems from the light-gathering limitations of the Minicam8M sensor structure at extremely fast focal ratios (F4), rather than from mechanical obs...