Unusual? ghosting pattern - Forums

Topic starter

Dec 11, 2023

After reducing last nights Ha data [26 x 600sec subs] of the Dolphin Head nebula, I was surprised to see the following pattern of ghosts caused by the bright star o 1 Can Majoris.

[This is the star-Xtermatined version]

The pattern is also there on all subs [with a meridian flip in the middle]. I don't think I have even seen such a regular rectangular pattern like this before, and I suspect it may have something to do with multiple reflections off the sensor window. THis may be quite a common occurance, but I have certainly never seen this before with a bright star [including with Alnitak].

Perhaps 600secs was too much for such a bright star - but I didn't see the same pattern in [OIII]. Has anyone encountered a similar ghosting phenomenon?

GSO RC8/Optolong Ha filter /ZWO ASI 6200MM

Helpful Insightful Respectful Engaging

Sean Mc

Dec 11, 2023

Ok I’m a noob but… how the heck do you even get a pattern like that?!?

Brian Boyle

Topic starter

Dec 11, 2023

Sean Mc:
Ok I’m a noob but… how the heck do you even get a pattern like that?!?

Hi Sean, I am less of a noob, but it still has me a little perplexed....

Wei-Hao Wang

Dec 12, 2023

I also got something similar recently. The telescope is a CDK and the filters are Ha, so kind of similar to your situation. The Ha filters are Baader 6.5 nm (top panel) and Chroma 5 nm (bottom panel).

Highly experienced people on Cloudy Nights told me it's reflection and diffraction of the micro lenses on the sensor and bouncing back from the filter. It can be more apparent on NB filter images, but I also got this in some of my broad-band filter images.

There is apparently no solution to this. A narrower NB filter may suppress it a little (like the bottom panel in my example image), but won't make it go away.

Helpful

Sean Mc

Dec 12, 2023

I would have thought that there would be more than 15 micro-lenses on the sensor. Doesn’t seem plausible to me ¯\_(ツ)_/¯

Brian Boyle

Topic starter

Dec 12, 2023

Thanks Wei-Hao, I don't know if I can say its good to see I am not alone.

Certainly reflections from the sensor window can give a rectlinear pattern e.g. the reflections from a laser when I set up to do a sensor tilt adjustment. This must be the internal reflection from the backside of the camera window back onto the sensor.

If so, I am surprised a) that I haven't seen it before b) more people haven't commented on it. It certainly makes the Dolphin Head a little challenging - unless there a good post-processing techniques to fix.

Although the reported instances are RC and CDK, since we are likely dealing with a sensor reflection due to a point source, I doubt it is limited to slower f-ratio scopes.

Helpful Insightful Respectful Engaging

Wei-Hao Wang

Dec 12, 2023

Hi Brian,

I think the diameter of the donuts/rings (plus the system F ratio) can be used to infer the separation between the two reflecting surfaces. In my case, a two-time reflection between the sensor and the filter can exactly explain the size of the ring. Because of this, I am sure the camera window does not play a role.

The spacing of the different donuts/rings is determined by diffraction. One can use lambda/D (where D is pixel size) to infer the diffraction angle. That multiplied by distance of the two reflection surfaces (multiplied by 2) will be the spacing between the rings. In my case, this exactly matches the pixel size of my camera.

I recall when I use TAK Epsilon or TOA to image Antares, I see this around Antares. On CN, this was asked from time to time. So many people definitely have seen this in their images before, but probably not everyone asks or complains about this.

Helpful Insightful

GalacticRAVE

Dec 12, 2023 · Edited: Dec 12, 2023

Brian,

these are reflections/refractions between the microlens array on the chip (which creates the grid) and the reflective window on the camera and/or the filters. People see that even with astrodon and chroma filters. There is a extensive discussion on cloudy nights (several threads)

E.g. https://www.cloudynights.com/topic/752319-what-are-these-artifactsreflections-antlia-35nmzwo-294mm/

with Mark Shelley (sharkmelley) providing some deeper insights by modelling.

matthias

ps I see it in optolong Ha and Sii with a 294mm (my recent IC417 picture was plagued by this), but also got an Baader Ha sent for testing, same story.

Helpful

Brian Boyle

Topic starter

Dec 12, 2023 · Edited: Dec 12, 2023

Matthias, Many thanks for pointing me in the right direction. The diameter of the halo means that the reflecting surfaces are 9mm apart which is either the approximately the sensor - camera window distance., or camera window to filter. It would make more sense it is were the latter since I don't see it with [OIII].

Irrespective of the source, it is clear that there is little I can do about it with hardware, so its another trick in post-processing I will need to learn. Or perhaps just avoid targets photobombed by bright stars….

CS Brian

Helpful Insightful Respectful

Wei-Hao Wang

Dec 12, 2023 · Edited: Dec 12, 2023

As for how to solve it, I think it either will be some dirty Photoshop work, or re-imaging the area with the star outside the FoV. I plan to do both.

For the source of the problem, I believe the sensor has to be involved. Otherwise there is no reason for the multiple reflections to favor the exact horizontal and vertical orientations with respect to the sensor.

Well Written Insightful Respectful Concise

Jens

Dec 12, 2023

Have you tried to put the camera onto a different scope? You should get something similar there as well. I've never seen such perfectly spaced reflections before.

Well Written

GalacticRAVE

Dec 12, 2023

Hi Brian,

Indeed, there is no obvious hardware fix as far as I can tell, even spenidng a fortune on filters seems not to help. Software wise, I used with reasonable success Vincent Peris multi-scale gradient method

https://pixinsight.com/tutorials/multiscale-gradient-correction/index.html

So take a wide field image (without artefacts), register it to the hires image, make both images star less. Then remove smaller scales with MLP, to only have the filter halos and the bright large scale nebulosity on the hires, and the large scale nebulosity on the low res. then substract the two images from each other with a scaling factor k such that the large scale nebulosity mostly disappears. Result is a smooth background image with the filters imprinted. subtract
the background from the hires image scaled by another scaling factor k such that the filter halos become as invisible as possible. Minor touchups then in
photoshop (the new AI fill works nicely), usually only necessary for the main filter image.

Matthias

Helpful Insightful Respectful

Brian Boyle

Topic starter

Dec 13, 2023 · Edited: Dec 13, 2023

Last posted my image after following Vincent Peris' tutorial. Thanks, Matthias!

Wei-Hao Wang

Dec 13, 2023

Hi Brian,

Do you have an H-alpha-only image to show how well it works?

Well Written Respectful

Wolfgang Promper

Dec 13, 2023

Hello Brian

I just saw that you are using a GSO RC, I had a 12" a couple of years ago.
It was a great telescope but it showed artefacts pretty simmilar to what you are seeing.
The problem was that the main mirror baffle is to short, at least it was on my scope.
I made an extension with cardboard and the reflections were gone, it might be worth to try that.
Wolfgang

Helpful Concise Supportive

Ryaja

Dec 13, 2023

If your using a reflector that could be images of the support structure showing up somehow.

Mark Ellis (PhoenixAstro)

Dec 13, 2023

Wei-Hao Wang:
I also got something similar recently. The telescope is a CDK and the filters are Ha, so kind of similar to your situation. The Ha filters are Baader 6.5 nm (top panel) and Chroma 5 nm (bottom panel).

Highly experienced people on Cloudy Nights told me it's reflection and diffraction of the micro lenses on the sensor and bouncing back from the filter. It can be more apparent on NB filter images, but I also got this in some of my broad-band filter images.

There is apparently no solution to this. A narrower NB filter may suppress it a little (like the bottom panel in my example image), but won't make it go away.

Humm Strange that you get this, I have a CDK 14 and I use Chroma 3nb filters and have never see this on any of my stuff. I agree that it's a reflection, maybe it's not bouncing back from the filter but maybe from the back element of the corrector in the CDK. I have see this sort of thing with my Edge 11" but not in a pattern such as this. Hope someone has a definitive answer to this mystery. I'll be keeping an eye out on this.

Brian Boyle

Topic starter

Dec 13, 2023

Wolfgang Promper:
Hello Brian

I just saw that you are using a GSO RC, I had a 12" a couple of years ago.
It was a great telescope but it showed artefacts pretty simmilar to what you are seeing.
The problem was that the main mirror baffle is to short, at least it was on my scope.
I made an extension with cardboard and the reflections were gone, it might be worth to try that.
Wolfgang

Hi Wolfgang,

Great point. And I have seen the youtube video about this issue. Fortunately, I have the version where GSO added and extension to the 1ary mirror baffle to eliminate this stray light problem. But that doesn't mean to say it works 100%....

However, this is the first only time I have seen this in over 3 years of imaging with this scope. The pattern and size strongly suggest an interaction between the sensor and camera window. If this is camera reflections, why does this shows up in only one filter?

CS Brian

Mark Ellis (PhoenixAstro)

Dec 13, 2023

Just out of curiosity, I pointed my CDK14 towards a bright star and took a few test shots. I did get a similar reflection, at least with my Lum filter. Camera is a QHY600M. Later I will do the same thing with my Refractor and see if I get something similar with that, It has a QHY 268M. Both setups have Chroma Unmounted 3nm filters.

Mark Ellis (PhoenixAstro)

Dec 13, 2023

Brian Boyle:
Wolfgang Promper:
Hello Brian

I just saw that you are using a GSO RC, I had a 12" a couple of years ago.
It was a great telescope but it showed artefacts pretty simmilar to what you are seeing.
The problem was that the main mirror baffle is to short, at least it was on my scope.
I made an extension with cardboard and the reflections were gone, it might be worth to try that.
Wolfgang

Hi Wolfgang,

Great point. And I have seen the youtube video about this issue. Fortunately, I have the version where GSO added and extension to the 1ary mirror baffle to eliminate this stray light problem. But that doesn't mean to say it works 100%....

However, this is the first only time I have seen this in over 3 years of imaging with this scope. The pattern and size strongly suggest an interaction between the sensor and camera window. If this is camera reflections, why does this shows up in only one filter?

CS Brian

That's a thought, I have a CDK14, I have see someplace that people have made 3d printed baffle extension. May be able to find them out on the interverse someplace.

Lorenzo Siciliano

Dec 13, 2023

Hi Brian.
Maybe I'm just saying something wrong, but… couldn't it be an IR leakage in the filter?
When I do pictures with a clear filter, I usually see around bright stars some odd patterns, similar to yours, that mostly disappear when I switch to an IR-cut (luminance) filter.
Maybe it's worth to investigate…
Ciao
Lorenzo

Helpful Insightful Respectful

Wei-Hao Wang

Dec 13, 2023

Hope someone has a definitive answer to this mystery.

To me, the answer cannot be more clear. The explanation offered by Mark on CN (see the link provided by Mathias) is very convincing and has predictive power. The pixel size of my camera, the spacing between the sensor and filter, and the focal ratio can exactly predict the size and pattern in my image. To me, this is a scientific proof that this is caused by diffraction of the pixel array on the sensor plus a two-time reflection between the sensor and the filter.

In general, the second reflection can occur at somewhere else in the optical path (like the optical window of the camera, in Brian's case), but the first reflection must be at the sensor surface. And because diffraction pattern size is proportional to wavelength and inverse-proportional to pixel size, if one uses a broad-band filter and a camera with somewhat large pixels, chance is good that the multiple reflected ghost images become blurred and blended. And if the star is not particularly bright, it may not get noticed.

Helpful Insightful Respectful

Mark Ellis (PhoenixAstro)

Dec 13, 2023

Wei-Hao Wang:
Hope someone has a definitive answer to this mystery.

To me, the answer cannot be more clear. The explanation offered by Mark on CN (see the link provided by Mathias) is very convincing and has predictive power. The pixel size of my camera, the spacing between the sensor and filter, and the focal ratio can exactly predict the size and pattern in my image. To me, this is a scientific proof that this is caused by diffraction of the pixel array on the sensor plus a two-time reflection between the sensor and the filter.

In general, the second reflection can occur at somewhere else in the optical path (like the optical window of the camera, in Brian's case), but the first reflection must be at the sensor surface. And because diffraction pattern size is proportional to wavelength and inverse-proportional to pixel size, if one uses a broad-band filter and a camera with somewhat large pixels, chance is good that the multiple reflected ghost images become blurred and blended. And if the star is not particularly bright, it may not get noticed.

This sounds like something the camera manufacturer's could work on if enticed enough to. Put some sort of anti-reflection coating on the inside of the window. Some filter manufacturer's are doing it with their filters so why not with the windows. Food for thought.

Wei-Hao Wang

Dec 14, 2023

We already know that even the best filters can cause this, but not always. Some camera windows can also cause this. So perhaps some astro cameras use very good AR-coated glass for the window, while other cameras use somewhat less good ones. Or maybe they are equally good in the common standard for AR coating, and it's just some subtle difference in the glass or coating formula that causes this. It's hard to tell now.