Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

New to narrowband imaging- question about strange pattern in bright star halo

  • Please log in to reply
72 replies to this topic

#51 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3822
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 06 February 2017 - 04:34 AM

 

 

As a side note...the ability to cancel reflections depends on there being additional reflecting photons at the same time that can cancel the original photons reflected. Maybe with low flux, that becomes more difficult? 

 

Now you're getting into interesting physics :)

 

My understanding (I may be wrong) is that a stream of single photons fired at a double slit can create diffraction patterns even though it is an interference process.  If so then I'm guessing the same is true of the interference process used by AR layers.

 

Mark

 

Light is behaving purely as a wave (a photon probability density wave/function if you like) in these matters and interferes much as you would expect using classical physics.



#52 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 15 February 2017 - 05:18 PM

I have also found this issue on the brightest stars. This are my first images with the new ASI1600 (january 2017) with the FSQ85 and LRGB filters. Gain 200 Offset 60. 60" for RGB and 90" for L.

 

This is a crop from a Crescent Nebula image.

 

Attached Thumbnails

  • IMG_3956.JPG

Edited by AstroCatinfo, 15 February 2017 - 05:18 PM.

  • sharkmelley likes this

#53 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 15 February 2017 - 05:20 PM

And another from the Horsehead:

Attached Thumbnails

  • IMG_3958.JPG


#54 sharkmelley

sharkmelley

    Vanguard

  • *****
  • Posts: 2244
  • Joined: 19 Feb 2013

Posted 15 February 2017 - 06:19 PM

I have also found this issue on the brightest stars. This are my first images with the new ASI1600 (january 2017) with the FSQ85 and LRGB filters. Gain 200 Offset 60. 60" for RGB and 90" for L.

 

This is a crop from a Crescent Nebula image.

That really is a good example of the effect! 

 

The coloured fringes are very obvious - it is clear that as the wavelength shortens from red to green the pattern shrinks.  You can see this directly on the coloured fringes and you can also see it you blink between the red and green channels. This is exactly what is predicted by a diffraction explanation.

 

For the Horsehead image, which filters did you use?

 

Mark


Edited by sharkmelley, 15 February 2017 - 06:25 PM.

  • AstroCatinfo likes this

#55 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 16 February 2017 - 03:17 AM

The Horsehead is an RGB image using much more subs from R. I also used R as Luminance. The filters are Baader LRGB 31mm unmounted.

 

Exact detail of both images (gain 200 Offset 60 -20ºC binx1 in ALL images):

 

Jellyfish image (the one that has a zoom on the star). LRGB image:

 

L: 119x90"

R: 20x60"

G: 19x60"

B: 19x60"

 

Horsehead. L=R. LRGB

 

R: 100x90"

G: 30x60"

B: 30x60"

 

-----------

 

How can I avoid this effect? Changing the GAIN/OFFSET values? I am a bit worried about it... just purchased the camera :/


Edited by AstroCatinfo, 16 February 2017 - 03:17 AM.


#56 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 16 February 2017 - 03:21 AM

Microlens diffraction issues are not uncommon. I know that the ubiquitous KAF-8300 has demonstrated the issue as well. Not everyone experiences it even with the KAF...I don't know why. I don't think that many manufacturers are putting multicoated sensor cover glass in their cameras, but that might be a solution. 

 

I'll mention the issue to Sam. The ZWO guys seem to like solving little problems, and it is very likely they would find a solution if they knew about the issue and took some time to investigate it. 

I don't doubt about the seriousness of Sam and their team. The have always given me great service and fast solutions.

 

Thanks Jon for this. I will also contact them.



#57 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 17 February 2017 - 05:23 PM

Any news on this issue?



#58 MoonPrince

MoonPrince

    Mariner 2

  • -----
  • Posts: 206
  • Joined: 27 Mar 2017

Posted 19 October 2017 - 04:46 AM

Thank you for all the insights everyone.

 

A few responses to things brought up that might help lead us in the right direction in determining why the pattern shows up with certain systems.

 



I'm wondering if maybe a slight bit of moisture on the window may be causing it, as I noticed this on my ASI1600 - and then it went away as the moisture eventually left?

I was not using any dew control on the camera, but plan to in the future, as I live in Delaware, which is a fairly humid state. Once I have this set up, I will try do comparisons.

 



Well, I think it is a combination of f-ration and aperture. I image at f/4 with a 150mm aperture...which is pretty darn fast for stars, since all the light from the star is concentrated into a point (or effectively so). When I encounter a bright star, it is BRIGHT, and clips heavily within seconds, even with a 3nm narrow band filter. I suspect that makes me more succeptible to the issue than say someone who was using an f/6 80mm refractor. 

I was imaging with an 80mm f/7.5 refractor, and with the 0.85x reducer, the effective focal ratio was about f/6.3. Quite a bit slower than your system, but as GeneralT001 pointed out, I was clipping stars, so maybe I should try again with shorter subs and see if I can see the pattern still. I am pretty sure it wasn't visible at 60 seconds as I did my framing and focusing, but I can't be positive.

 



A conclusive test requires images taken with 2 narrowband filters e.g. Ha and Oiii. Does the pattern spacing then change with wavelength as predicted by a diffraction model?

This reminded me that I did do a sample OIII image the other night. It is the same camera settings (Unity, 5 min. sub) and same focus as the Ha.  One thing that throws off the comparison is my Ha filter is the 5nm bandpass and the OIII is 3nm.  Anyways, the pattern is clearly visible in the Ha, and not visible in the OIII, but I figure that just comes down to the 3nm OIII not letting in as much light as the 5nm Ha. So, not conclusive, but still interesting.

b3AKRMMh.png

Astrodon H-alpha 5nm

VcgrBMKh.png

Astrodon OIII 3nm

Does it means, the halo is cause by filters not the camera? could you use this camera without filters and shoot a bright star, then take a look if halo is still there:)



#59 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8696
  • Joined: 12 Oct 2007

Posted 19 October 2017 - 05:51 AM

The simulation that I did in this thread:

 

https://www.cloudyni...act-w-canon-6d/

 

shows a good match to the effect in a 6D with no assumptions other than coherent waves being emitted by an array of point sources spaced at the pixel spacing and forming an image very close by on a screen.  In that case the distance was about 1mm, corresponding to a reflection off the sensor cover slip about 0.5mm thick.

 

It assumes the beam is focused to a circular patch on the detector and forms an image some distance away.  The rest is just brute force calculation of the diffraction pattern with no assumptions of near field, far field - etc.

 

So - these effects are likely due to a diffraction pattern that has traveled a very short distance in one bounce up and down back to the pixels - and if the filter or camera window were involved, the throw distance would be much larger and the effect would be very different.

 

I think the difference in Ha and Oiii here and elsewhere is due to different reflectance values off the microlenses, among other things.  If the lenses or cover slip have a simple "V" anti-reflection coating, they would have minimum reflectance in green and higher in red and blue.  So it is at least consistent.

 

I do see the effect in some of my asi1600 images but it is rare.  I haven't actually modeled my own pattern yet.

 

Frank



#60 2ghouls

2ghouls

    Viking 1

  • *****
  • topic starter
  • Posts: 702
  • Joined: 19 Sep 2016

Posted 19 October 2017 - 09:19 AM

Does it means, the halo is cause by filters not the camera? could you use this camera without filters and shoot a bright star, then take a look if halo is still there:)


No, the astrodon filters are fine. I’ve used them with other optics and same camera and not had problems. Later in the thread people point out that I was out of focus on the o3 test and once I was in focus the pattern does show up just fainter in the o3.
It seems to me to be something about the combo of the ZWO ASI 1600mm with certain optical systems that causes it. Personally, I do not think it is related to f ratio as I have never seen the pattern with my 200mm, f2.8 lens even when shooting the Sadr region (Sadr is very bright compared to the surrounding nebulosity). I’ve also never seen it with a 94mm f/7 vernonscope “Brandon” triplet, but I haven’t done rigorous testing. In any case, it makes me want to sell my 80mm doublet with reducer/corrector, as the pattern is annoying to me and I think that’s why I’ve been subconsciously avoiding using that scope with this camera.


Sent from my iPhone using Tapatalk

#61 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8696
  • Joined: 12 Oct 2007

Posted 19 October 2017 - 03:56 PM

A separate factor could be the actual size of the star spot, in pixels, on the sensor.  This will involve a combination of seeing, f/ratio, focus - and more.  The spot needs to be big enough to illuminate a sufficient number of pixels to generate a visible diffraction pattern.

 

In my case I see it with my EdgeHD11 at f/10, but a recent image with Canon 135 at f/4, which is a very sharp and short focus lens - did not show it, even with a very bright star in the field.

 

So I suspect it is more correlated with longer focal length systems where the star spot is seeing limited and spans 5-10 pixels or so.

 

Defocus would make the star spot bigger - but also fainter - so it would need a correspondingly longer exposure.

 

I think the only role a filter would play is in the way the relevant surfaces are AR coated - if they are.  The two surfaces are 1)  the microlenses and 2)  The sensor cover slip.

 

Frank


  • 2ghouls likes this

#62 mistateo

mistateo

    Apollo

  • *****
  • Posts: 1238
  • Joined: 02 Feb 2017
  • Loc: San Diego, CA

Posted 20 November 2017 - 05:11 AM

For what it's worth, I get this issue with my ASI1600MM-C as well.  My ZWO LRGB set shows the diffraction pattern on bright stars, the ZWO narrowband set the effect seems quite a bit less, but they are prone to halos with both my scopes.  The Orion Extra Narrowband set eliminated the halos (except the Oiii, must be a bad sample, was tempted to throw it in the garbage as it was the worst filter I have ever seen, but decided to try an exchange), but the diffraction pattern is definitely there with them.

 

Happens on both of these setups:

 

Orion Eon 120mm w/.85x reducer/flattener, 765mm focal length f/6.375

Stellarvue SB080T-25FT w/.8x reducer/flattener, 384mm focal length f/4.8



#63 stevekaczmar

stevekaczmar

    Sputnik

  • -----
  • Posts: 26
  • Joined: 18 Dec 2015

Posted 20 November 2017 - 11:18 AM

Seeing that this thread is still going I could use an opinion as to why this "reflection" {if that is what it is} all of a sudden shows up. The center red circle is where the "spot was while shooting the Sisters..I moved to the Horsehead and the spot moved to the larger red circle.At the sister it was perfectly round and was like a double halo effect.Once moved to the Horse it became oblong with larger halo..Setup..70mm stellarview with reduce.ZWO178 and an orion light pollution filter..I had been using the same setup for a while now and this just started to show up,just like that..lol..I cleaned,checked,recleaned ..long story short after removing the reducer the problem went away.So why all of a sudden did these "spots" start showing up?I have 2 theories,1 the leaves from the trees are gone now and could it be picking up a reflection from the neighbors vapor light that was blocked by the tree before? 2. the reducer has a problem? I leave my gear in my roll off shed,no heat provided.Is it possible the metal of the reducer has either gotten to hot or to cold to cause the metal to expand or contract to cause this? The telly its self may have be out of alignment somehow so when using the reducer it is causing this spot? It is not a scratch or dirt plus the spot moved to another area when I move to another target leading me to beleive it is a light reflection. But using the LP filter that should not happen,right? Any ideas people? Thanks..

 

weird reflectionV

Edited by stevekaczmar, 20 November 2017 - 11:43 AM.


#64 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 19 December 2017 - 06:03 PM

I have upgraded to ASI1600MM-C PRO so I will be able to do a comparision and see if the effect of the paterns remains. Stay tuned the following weeks wink.gif


Edited by AstroCatinfo, 19 December 2017 - 06:03 PM.

  • 2ghouls likes this

#65 mistateo

mistateo

    Apollo

  • *****
  • Posts: 1238
  • Joined: 02 Feb 2017
  • Loc: San Diego, CA

Posted 27 December 2017 - 05:46 PM

The pro version uses the same panasonic sensor.  The issue won't go away in the new camera.  Panasonic has already stated they are not interested in applying antireflective coatings to the sensor protective glass element of this sensor.  Their primary application of the sensor is night time security cameras, astrophotography applications only make up a small fraction of the sales on these sensors.  Microlens diffraction is only a problem on pretty bright stars, and can be minimized by only exposing to 20x your read noise and taking more subs.  I am disappointed by this trait of the camera as well, but am learning to live with it.  Such a shame I can't get a great photo of the horsehead and flame nebula composition using this camera, but many targets work just fine.


  • AstroCatinfo likes this

#66 sharkmelley

sharkmelley

    Vanguard

  • *****
  • Posts: 2244
  • Joined: 19 Feb 2013

Posted 28 December 2017 - 03:16 AM

Microlens diffraction is only a problem on pretty bright stars, and can be minimized by only exposing to 20x your read noise and taking more subs. 

Really?  I would be interested to see the evidence for this. 

 

If in both cases you are receiving the same number of total photons then surely the same number of photons contribute to diffraction effects in both cases.  Otherwise there is some very interesting physics going on.

 

Mark


  • happylimpet likes this

#67 vlad dumitrescu

vlad dumitrescu

    Explorer 1

  • -----
  • Posts: 70
  • Joined: 08 Sep 2008
  • Loc: București

Posted 28 December 2017 - 04:54 AM

Ok, so i recently got this camera and it seems like i can t use it to shoot the horsehead. I m so dissapointed :( Below a STF stretch of a 20x300s stack shot on my FSQ85 and optolong h-alpha 7nm filter. Is it really that hopeless?

Vladmicshot.jpg


  • AstroCatinfo likes this

#68 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8696
  • Joined: 12 Oct 2007

Posted 28 December 2017 - 05:59 AM

A 3nm Ha filter will greatly reduce the starlight while letting the nebulosity in at full strength.  Aside from that - I think all you can do is look for different compositions that don't have the star in view.  Horsehead images don't need to include that star.

 

Frank


  • AstroCatinfo likes this

#69 mistateo

mistateo

    Apollo

  • *****
  • Posts: 1238
  • Joined: 02 Feb 2017
  • Loc: San Diego, CA

Posted 28 December 2017 - 04:39 PM

 

Microlens diffraction is only a problem on pretty bright stars, and can be minimized by only exposing to 20x your read noise and taking more subs. 

Really?  I would be interested to see the evidence for this. 

 

If in both cases you are receiving the same number of total photons then surely the same number of photons contribute to diffraction effects in both cases.  Otherwise there is some very interesting physics going on.

 

Mark

 

I suppose it is possible that my two data sets weren't processed exactly the same, but I shot m45 once with 3 minute LRGB subs and once at L 10s, R & B 30s, and G at 20s to give around 20x read noise.  Both of them were shot at unity gain with the same total integration time, and in the image with longer subs the main bright stars looked like snow flakes, while the other had what looked like 4 chunky diffraction spikes in a cross pattern, and then 4 smaller ones, rotated 45 degrees.  I will try to dig up the example images when I get home.  I wonder if it is possible that the longer subs are more likely to not only increase the FWHM a little bit, but possibly bloom light across more pixels.  Not sure if blooming is an issue with CMOS cameras or not, or if that would contribute to the intensity of the microlens effect in the image itself.



#70 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23384
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 28 December 2017 - 05:16 PM

Ok, so i recently got this camera and it seems like i can t use it to shoot the horsehead. I m so dissapointed frown.gif Below a STF stretch of a 20x300s stack shot on my FSQ85 and optolong h-alpha 7nm filter. Is it really that hopeless?

Vladmicshot.jpg

There are a few specific objects that are problematic with the ASI600 due to it's microlens diffraction. Well, more specifically, a few specific stars. Alnitak is enemy #1, basically. And this is not just the case with the ASI1600, although it tends to be a bit worse due to the microlens reflection, but it is an issue with scopes that have higher blue dispersion (you end up with a bright blue cast over most of the image in color combinations), scopes that have big fast apertures (you usually get obvious aperture-shaped reflections), etc. Alnitak is THE star that annoys a majority of imagers with a very wide range of equipment. A close second is Gamma Cas, another very bright star in Cassiopeia near the Ghost of Cass nebula. It, too, is ultra bright and has been a problem for many imagers with a wide range of equipment. It just gets a bit worse with the ASI1600 due to the microlens diffraction. There are some other stars out there that are similarly problematic, although usually not to the degree of these two, however it is a minority of objects that are affected.

 

As with any scope that has problems with Alnitak or Gamma Cass, one solution is to orient the image such that the problematic star is out of the frame. With reflection and dispersion issues, this is only partially successful since a reflection can still occur off of other elements in the train, and can still be a problem so long as the star is within the imaging circle (which is sometimes many times the area of the sensor itself). With the ASI1600, all you need is for the star to be JUST off the frame, and the issue will disappear entirely. So that gives you more options.

 

The interesting thing about this effect is it is actually now well  understood thanks to the knowledge and experience of guys like Mark (sharkmelly) and Frank (freestar8n). I think Frank was actually working on a tool that could be used to cancel or at least mitigate the effect of the diffraction due to the knowledge those two gained while learning about the effect when it occurred with other kinds of cameras (no, the ASI1600 is not the only camera to experience the effect!) It seems it can be well-modeled, which means at some point there may be a solution to the problem. (Food for thought).


  • stevekaczmar and AstroCatinfo like this

#71 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8696
  • Joined: 12 Oct 2007

Posted 28 December 2017 - 06:07 PM

Hi Jon-

 

No - I was just modeling it to make sure the pattern of spots matched the coverslip thickness and the overall shape agreed well - and I think it does.  The exact pattern depends on how light reflects off the microlenses and would be very hard to model exactly - in terms of how each spot in the pattern appears.  Sometimes they are fairly smooth and other times they are somewhat donut.  And if the central star saturates then it is really hopeless.

 

But you could try placing a bright star in about the same place as Alnitak is in the image and expose it - to create a good match to the pattern that maybe you could subtract with careful processing.

 

One thing that could help is that the pattern is largely set by the size of the star spot on the pixels and the f/number - but how visible the pattern is will depend heavily on aperture and exposure time.  At the same time, the nebulosity in the image will just depend on f/number.  I did some imaging with a Canon 135 f/2 stopped down to f/4 with Antares in the field - and there was no sign of the pattern, even though Antares was blazing.  But that is an aperture of only about 30mm compared to 280 for my EdgeHD11 - or a factor of about 90 in brightness of the pattern for a given exposure time.  This may also play a role in making the problem less of an issue for dslr's, which is what the sensor is for in the first place.  I know that dslr's do show the pattern occasionally - but I don't think it's a common problem - which may be why they don't bother coating them.

 

I think we are in a transition time regarding what sensors we have to work with and at what cost so I hope these issues go away.  But in the meantime it's good to know what you can and can't do about it.  I see the pattern in some of my images and it is faint enough that it doesn't bother me.  But I agree for the OP's horsehead image it is distracting.  But there are ways to compose the horsehead in portrait mode to exclude Alnitak - and then you get a strong contrast of the extended nebulosity going up vs. the darker area below.  Alternatively, with some processing effort you may be able to create the pattern by itself and subtract it.

 

Frank



#72 GeneralT001

GeneralT001

    Vanguard

  • *****
  • Posts: 2435
  • Joined: 06 Feb 2012
  • Loc: Nanaimo, BC

Posted 28 December 2017 - 08:51 PM

Too bad there is no way to just turn off the pixels that cover these stars and then just paint in a proper looking image of the star afterwards.



#73 AstroCatinfo

AstroCatinfo

    Ranger 4

  • *****
  • Posts: 350
  • Joined: 22 Jan 2016
  • Loc: Prades, Catalunya - Spain

Posted 06 January 2018 - 05:37 PM

The pro version uses the same panasonic sensor.  The issue won't go away in the new camera.  Panasonic has already stated they are not interested in applying antireflective coatings to the sensor protective glass element of this sensor.  Their primary application of the sensor is night time security cameras, astrophotography applications only make up a small fraction of the sales on these sensors.  Microlens diffraction is only a problem on pretty bright stars, and can be minimized by only exposing to 20x your read noise and taking more subs.  I am disappointed by this trait of the camera as well, but am learning to live with it.  Such a shame I can't get a great photo of the horsehead and flame nebula composition using this camera, but many targets work just fine.

You are right! Reflections remain there. I have tryied many ways to take the RGB of the Horse and Flame but reflections in G and B were always there... I think I will stop with the Ha image. New targets to come...




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics