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

Guided subs moving during long image session

  • Please log in to reply
25 replies to this topic

#1 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 08:40 AM

Hello

I am not sure is this question goes in this forum but it seems like it could belong to more than one.

 

Last night, I was doing OIII images of SH2-129 (Giant squid nebula). My exposures were 8 and 10 min to a total of 7h. I noticed that there is a shift between subs and it is more obvious when compare for example frame 1 vs frame 8 (about 40 min later). My stars were round and the total guided RMS about 0.7". Any idea what can cause that?

Thank you so much

 

Rodrigo

 

 

 

 



#2 Der_Pit

Der_Pit

    Mariner 2

  • -----
  • Posts: 257
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 14 September 2018 - 08:44 AM

As you don't tell anything about your equipment only some blind shots:

Flexure between guide scope and main scope, lose mounting of guide cam on OAG, non-perfect polar alignment causing image rotation.


  • jdupton and happylimpet like this

#3 Zebenelgenubi

Zebenelgenubi

    Mariner 2

  • *****
  • Posts: 246
  • Joined: 06 Apr 2012
  • Loc: Texas, USA

Posted 14 September 2018 - 08:57 AM

I saw a similar effect on long DSO exposures last September. The star images looked okay but the stacking correction got progressively larger.  Also the scatter gram in PHD was a tight grouping but with a definite bias from zero. It turned out that I had forgotten that I set my mount at "Solar Rate" for the eclipse and hadn't reset to 'Sidereal".


  • happylimpet likes this

#4 happylimpet

happylimpet

    Mercury-Atlas

  • *****
  • Posts: 2998
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 14 September 2018 - 09:00 AM

As you don't tell anything about your equipment only some blind shots:

Flexure between guide scope and main scope, lose mounting of guide cam on OAG, non-perfect polar alignment causing image rotation.

Agreed. My money's on PA.

 

I saw a similar effect on long DSO exposures last September. The star images looked okay but the stacking correction got progressively larger.  Also the scatter gram in PHD was a tight grouping but with a definite bias from zero. It turned out that I had forgotten that I set my mount at "Solar Rate" for the eclipse and hadn't reset to 'Sidereal".

Haha - I did the same with lunar rate a few months ago.



#5 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 09:05 AM

As you don't tell anything about your equipment only some blind shots:

Flexure between guide scope and main scope, lose mounting of guide cam on OAG, non-perfect polar alignment causing image rotation.

Sorry, I should have been more specific:

 

My equipment is:

 

Mount: Losmandy G11 (spring loaded RA), image telescope is WO 98mm, my guiding camera is the Orion package (50mm finder Orion SSAU)

My image camera is ASI 1600mm-C with filter wheel, using OIII. I use Pole master to do my polar alignment

 

Thank you



#6 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 09:09 AM

With non perfect PA, does the guided image still look Ok although there is a shift between subs?

 

Thank you

 

Rodrigo



#7 Der_Pit

Der_Pit

    Mariner 2

  • -----
  • Posts: 257
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 14 September 2018 - 09:14 AM

OK, so a reliable mount and supposedly good PA due to pole master.

Then either the mount of the finder scope is not stiff enough, or the missmatch in resolution is too large (if the guide cam has 20 arcsec/pixel it has a hard time guiding 2 arcsec accuracy...)

 

Yes, with PA error you will get a perfect guide curve, but still errors in the long exposure.

If you have taken several subs, just blink the first and the last.  Image rotation is easy to see.  If it's only shift it is rather the guider mounting.



#8 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 09:29 AM

OK, so a reliable mount and supposedly good PA due to pole master.

Then either the mount of the finder scope is not stiff enough, or the missmatch in resolution is too large (if the guide cam has 20 arcsec/pixel it has a hard time guiding 2 arcsec accuracy...)

 

Yes, with PA error you will get a perfect guide curve, but still errors in the long exposure.

If you have taken several subs, just blink the first and the last.  Image rotation is easy to see.  If it's only shift it is rather the guider mounting.

Thank you so much, I will check that tonight since I can do that with almost 60 subs I have



#9 bobzeq25

bobzeq25

    Voyager 1

  • *****
  • Posts: 13465
  • Joined: 27 Oct 2014

Posted 14 September 2018 - 09:56 AM

I used to see this.  As my PA and guiding have improved, it's gone away.



#10 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 09:58 AM

Do you think that a problem with the DEC balance can cause that too? My DEC is not spring loaded so I see a lot of correction sometimes



#11 jdupton

jdupton

    Apollo

  • *****
  • Posts: 1350
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 14 September 2018 - 10:20 AM

RRod,

 

Sorry, I should have been more specific:

 

My equipment is:

 

Mount: Losmandy G11 (spring loaded RA), image telescope is WO 98mm, my guiding camera is the Orion package (50mm finder Orion SSAU)

My image camera is ASI 1600mm-C with filter wheel, using OIII. I use Pole master to do my polar alignment

 

Thank you

 

   With a separate guide scope, good PA, and good guiding, then the problem is most likely differential flexure. The optical axis of the guide scope is moving with respect to the optical axis of the imaging telescope over time. The amount of this movement is so small, it is imperceptible to a mere mortal.

 

   There is a method that can help determine whether the slow image shift is being caused by flexure or by polar misalignment. Check the following in your raw images:

  • Stack all of your raw sub-exposures without first aligning them.
     
  • Closely examine the resulting stacked image.
    The stacking will show all stars as streaks in the image since they were drifting from one sub-exposure to the next.

    Case A: The streaks are slightly curved and your best estimate of the center of that curvature is in the area of the guide star you used (whether in the main imaged frame or just outside of it).
    Case A Conclusion: Polar alignment is not as good as it should have been. You are seeing field rotation caused by the polar alignment error. The whole image will tend to rotate around the guide star as the session progresses.

    Case B: The streaks are in mostly straight lines all of the same length and pointed along the same direction.
    Case B Conclusion: You have differential flexure. Your guide scope and imaging scope are not pointing the same direction as the night goes on. As the telescope tracks, gravity pulls on one slightly differently than the other.

 

   The amount this flexure can be exceedingly small. A shift of the optical axis of the guide scope with respect to the imaging scope as small as 0.001"  measured at the end of the guide scope will result in an error of about 1.27 arc-seconds of movement. Most folks are unable to visually detect 0.001" of flex over a period of several minutes but it adds up over a period of 7 hours of total exposure.

 

   If flexure is at the root of your drifting frames, you can determine the actual amount of flexure by measuring the total length in pixels of any straight lines in the above unaligned stacked image, multiplying by your image scale in arc-seconds per pixel, and then dividing by the total (7 hours) exposure time in minutes. That will give you the drift between the two optical axes in arc-seconds per minute. It will likely be a small number but it all adds up over a full night of imaging.

 

 

John


  • rgsalinger, psandelle and happylimpet like this

#12 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 10:31 AM

RRod,

 

 

   With a separate guide scope, good PA, and good guiding, then the problem is most likely differential flexure. The optical axis of the guide scope is moving with respect to the optical axis of the imaging telescope over time. The amount of this movement is so small, it is imperceptible to a mere mortal.

 

   There is a method that can help determine whether the slow image shift is being caused by flexure or by polar misalignment. Check the following in your raw images:

  • Stack all of your raw sub-exposures without first aligning them.
     
  • Closely examine the resulting stacked image.
    The stacking will show all stars as streaks in the image since they were drifting from one sub-exposure to the next.

    Case A: The streaks are slightly curved and your best estimate of the center of that curvature is in the area of the guide star you used (whether in the main imaged frame or just outside of it).
    Case A Conclusion: Polar alignment is not as good as it should have been. You are seeing field rotation caused by the polar alignment error. The whole image will tend to rotate around the guide star as the session progresses.

    Case B: The streaks are in mostly straight lines all of the same length and pointed along the same direction.
    Case B Conclusion: You have differential flexure. Your guide scope and imaging scope are not pointing the same direction as the night goes on. As the telescope tracks, gravity pulls on one slightly differently than the other.

 

   The amount this flexure can be exceedingly small. A shift of the optical axis of the guide scope with respect to the imaging scope as small as 0.001"  measured at the end of the guide scope will result in an error of about 1.27 arc-seconds of movement. Most folks are unable to visually detect 0.001" of flex over a period of several minutes but it adds up over a period of 7 hours of total exposure.

 

   If flexure is at the root of your drifting frames, you can determine the actual amount of flexure by measuring the total length in pixels of any straight lines in the above unaligned stacked image, multiplying by your image scale in arc-seconds per pixel, and then dividing by the total (7 hours) exposure time in minutes. That will give you the drift between the two optical axes in arc-seconds per minute. It will likely be a small number but it all adds up over a full night of imaging.

 

 

John

Thank you, Just a question about possible flexure, I have my guiding scope aligned to my image scope, so when I choose the star, I make sure that is also in my main scope field of view, would that take care of flexure?



#13 jdupton

jdupton

    Apollo

  • *****
  • Posts: 1350
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 14 September 2018 - 10:40 AM

RRod,

 

   No, unfortunately not. Regardless of how the optical axes are initially aligned, the two can be affected differently by gravity as the mount tracks across the sky. 

 

   There are a large number of posts here on CloudyNights regarding common causes of differential flexure and approaches to taming it. Your's doesn't sound like an extreme case since you see it in sub-to-sub movement but your stars are still acceptably round. That is what counts most -- nice round stars.

 

   Some of the more common contributors to flexure are dangling cables, tall mounting rings on the finder/guider, and sometimes plastic mounting elements. Show us a photo of you imaging rig and folks can point out points of improvement. 

 

   In my opinion, if you are getting good star shapes, nothing needs to be done unless you want to try to improve the drift on theoretical grounds.

 

 

John


Edited by jdupton, 14 September 2018 - 10:41 AM.

  • psandelle and happylimpet like this

#14 CygnusBob

CygnusBob

    Sputnik

  • -----
  • Posts: 49
  • Joined: 30 Jun 2018
  • Loc: Las Vegas, NV

Posted 14 September 2018 - 10:56 AM

Another issue is atmospheric refraction.  Even with perfect polar alignment, over 7 hours of guiding, atmospheric refraction will result in a certain amount of drift.  At zenith this error source is zero. However, at an elevation angle of 45 degrees there is about 60 arc seconds of angle error.



#15 cfosterstars

cfosterstars

    Surveyor 1

  • *****
  • Posts: 1649
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 14 September 2018 - 12:33 PM

Dither?



#16 Dan Crowson

Dan Crowson

    Surveyor 1

  • *****
  • Moderators
  • Posts: 1856
  • Joined: 08 Oct 2010
  • Loc: O'Fallon, MO

Posted 14 September 2018 - 12:36 PM

This sounds like classic flexure. There used to be an article in the DeepSkyStacker wiki that would even explain the procedure to determine it.

Dan



#17 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 12:51 PM

Thank you for the information, considering this movement i does dither is still necessary?

#18 choward94002

choward94002

    Mariner 2

  • *****
  • Posts: 203
  • Joined: 25 Jul 2016

Posted 14 September 2018 - 01:02 PM

Dithering solves the problem of hot pixels which you haven't mentioned ... right now you've got image shifts, which is either bad PA alignment (image rotation) or flexure in your imaging path ... both of which can be determined by looking at the first, middle and last exposures (as others have mentioned).  I use AstroBlinker [https://www.cloudyni...o-blink-images/] to find those kinds of problems here, you will be able to see quick quickly what's happening with this ...

 

Clear skies!


Edited by choward94002, 14 September 2018 - 01:04 PM.


#19 Dan Crowson

Dan Crowson

    Surveyor 1

  • *****
  • Moderators
  • Posts: 1856
  • Joined: 08 Oct 2010
  • Loc: O'Fallon, MO

Posted 14 September 2018 - 01:06 PM

You will want to dither but (with the assumption this is flexure), it is all rather dithering for you. :/ Make sure everything is tight. I had the same kind of issues with three different setups. I was getting roughly 6 pixels in 10 minutes. An off-axis-guider will eliminate it for sure but comes with other complications.

Dan



#20 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 01:27 PM

Dithering solves the problem of hot pixels which you haven't mentioned ... right now you've got image shifts, which is either bad PA alignment (image rotation) or flexure in your imaging path ... both of which can be determined by looking at the first, middle and last exposures (as others have mentioned).  I use AstroBlinker [https://www.cloudyni...o-blink-images/] to find those kinds of problems here, you will be able to see quick quickly what's happening with this ...

 

Clear skies!

Thank you, that is a cool program, I will give it a try



#21 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 14 September 2018 - 01:29 PM

You will want to dither but (with the assumption this is flexure), it is all rather dithering for you. :/ Make sure everything is tight. I had the same kind of issues with three different setups. I was getting roughly 6 pixels in 10 minutes. An off-axis-guider will eliminate it for sure but comes with other complications.

Dan

I thought about and OAG and I bought one but the entire setup gets too heavy for my focuser. So I am thinking to use it when I use the OSC camera.



#22 Der_Pit

Der_Pit

    Mariner 2

  • -----
  • Posts: 257
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 15 September 2018 - 08:14 AM

Whether you can use the drift/rotation as dither (it's in principle somewhat comparable to RA-only dithering) will mostly depend on the characteristics of your imaging sensor.  Fixed pattern noise plus a linear drift can lead to streaking artifacts ('rain') in the summed image.  That said, in the past for me this has worked quite OK, too.  As you also have the ASI1600MM: In doubt just take shorted subs, but more of them.  John already mentioned: The first important thing is that the single subs have (perfect wink.gif ) round stars.  



#23 bobzeq25

bobzeq25

    Voyager 1

  • *****
  • Posts: 13465
  • Joined: 27 Oct 2014

Posted 15 September 2018 - 09:56 AM

As you can see, there's a lot to consider.  Basic facts, hopefully they simplify.

 

Flexure is a common problem, the only solutions are to mount the guidescope very rigidly, or use an off axis guider.

 

Dithering is a useful technique to compensate for some camera noise issues.  It's best done carefully and systematically.

 

Focusers are also a common problem.  Now the solutions are a better scope or an aftermarket focuser.

 

What you're experiencing are things we all experience.  There are good solutions, but they require knowledge, care, and often, some money.  <smile>  There is no magic here.

 

Reread my post #9 in this light.


Edited by bobzeq25, 15 September 2018 - 09:59 AM.


#24 RRod

RRod

    Mariner 2

  • -----
  • topic starter
  • Posts: 224
  • Joined: 07 Aug 2006
  • Loc: Green Bay, WI

Posted 15 September 2018 - 12:28 PM

Thank you so much for all the advice and information 

Rodrigo


Edited by RRod, 15 September 2018 - 12:29 PM.


#25 tolgagumus

tolgagumus

    Vendor

  • *****
  • Vendors
  • Posts: 3455
  • Joined: 21 Jul 2014
  • Loc: NJ

Posted 16 September 2018 - 12:40 PM

With non perfect PA, does the guided image still look Ok although there is a shift between subs?

 

Thank you

 

Rodrigo

So it depends on the type of shift. With non perfect PA, you will get rotation between the first and the last sub even if you are guiding. Depending on how bad the misalignment is, it may not even be noticeable between two consecutive subs.  And the rotation will be centered around the guide star. 




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