26 replies to this topic

[axlns]

Hi everyone!

 

First of all, I'm newbie and own telescope only for a few months and had only several sessions taking pictures of DSO.

And long story short - I cant get round stars no matter what I tried.

 

Initially, I thought it might be guiding issue, bought guiding camera and now I'm guiding when taking pictures but still cant get round stars.

I also bought coma corrector (Baader MPCC III), that did not fix issues I'm having too.

And I spent a lot of time on collimating the scope and I believe it is well collimated right now (both laser collimator and cheshir showing that collimation is fine)

 

I think I already tried everything, so I decided to ask for help on this forum. Hopefully, you guys can point me in the right direction.

 

My setup

•     200mm f/5 Newtonian reflector (SkyWatcher)
•     SkyWathcher HEQ5 GoTo mount
•     Finder scope with guiding camera (QHY5L-II-M)
•     Sony Alpha 99 as my main imaging camera (its full-frame camera), connected to the focuser through Baader MPPC III to correct coma issues

 

The problem is that I cant really get round stars. They are more or less round in the center of frames, but not on the edges of frames.

 

And I believe there is a pattern in the way how stars are "elongated". I'm hoping it will give experienced people a clue to the reason of the problem.

The best way I can describe the pattern is that the stars are following in "circle direction". It's like when you are making a photo of the sky pointing to Polaris without guiding, you then see how other stars around Polaris are running in circles. Hopefully, I'm explaining it clearly. Also, I must point that the severity of "elongation" does NOT seems to depend on the length of exposure, i.e., 5seconds long exposures and 60seconds long seems to show almost the same level of elongation. Usually, I'm taking light frames with 60 seconds long exposure.

 

Here you can download the selection of RAW files if you like to closely inspect them (some are 5 seconds, some 60 seconds, note that its 183mb 7zip archive) - https://www.sugarsyn...09685078_662175

 

Let me show you a few screenshots of the light frame with M31 with its core in the center of the frame. Its 5 seconds exposure frame at 8000 ISO.

 

The central part (with more or less rounded stars) - https://prnt.sc/p75w1r

Top Left corner - https://prnt.sc/p75wh8

Top Mid side - https://prnt.sc/p75wqj

Top Right corner - https://prnt.sc/p75wy4

Mid Right side - https://prnt.sc/p75x5d

Bottom Right corner - https://prnt.sc/p75xel

Bottom Mid side - https://prnt.sc/p75xm9

Bottom Left corner - https://prnt.sc/p75y38

Mid Left side - https://prnt.sc/p75ycl

 

and just for the comparison - the central part of 60sec long exposure - https://prnt.sc/p7605u - the stars become "longer" and have "tails" and look somewhat like triangle - this is another problem i'm fighting too (stars as triangles) and cant get rid of it.

and bottom left corner of 60sec long exposure - https://prnt.sc/p761e2 - its very similar elongation to 5sec long exposure (same corner) - https://prnt.sc/p75y38   

 

Can you guys suggest what should I try to remove this elongation and get round stars?

 

Could it be guiding issue? Would it look like stars "running in circles" around the central object if the problems are in guiding?

I believe I have an issue with guiding in RA, but I'm not sure how this can lead to stars "running in circles", but maybe I dont understand something.

 

I believe I have more or less good DEC guiding but I'm aways hawing "saw-like" guiding in RA - see the screenshot of the guiding log  - https://prnt.sc/p75yps

Also maybe someone can detect issues in calibration - this is a screenshot of calibration log https://prnt.sc/p75z4w

I'm going to attach PHD2 guiding log and calibration log as files too; maybe someone can find some wrong things within them.

 

Or maybe this is a collimation issue?

Or the polar alignment issue?

Maybe balancing issue?

Or something entirely different?

 

I'm out of ideas and its really disappointing when you are wasting those rare clear sky nights and don't end up with pictures with decent round stars. I want to get it fixed and hoping you can help me. Thanks in advance!

Attached Files

•  PHD2_GuideLog_2019-09-16_213212.txt   27.58KB   18 downloads
•  PHD2_GuideLog_2019-09-17_022835.txt   84.38KB   9 downloads

Edited by axlns, 17 September 2019 - 08:26 AM.

[Galaxyhunter]

To me, it looks like field rotation.

[AhBok]

What method are you using for polar alignment?

[Jeff_Richards]

Is your camera at the required 55mm +/-1? With reducers and/or correctors it is essential that the imaging chip is at the specified distance from the corrector. You are using a full frame sensor so are you using the 48mm adapter with the necessary spacer per the Baader instructions?

[axlns]

Thanks for all your answers!

 

Is your camera at the required 55mm +/-1? With reducers and/or correctors it is essential that the imaging chip is at the specified distance from the corrector. You are using a full frame sensor so are you using the 48mm adapter with the necessary spacer per the Baader instructions?

 

Yes, I double checked, and back focus is 55mm with T-ring I'm using. 

I must add that Im still using M42 T-ring, I ordered M48 T-ring from local store and it did not fit well Baader MPCC III (looks like Baader's M48x0.75 thread actually not 0.75, that was annoying finding), so I ordered another M48 T-ring from Telescope-Express. and I have 3mm extension ring, so it still will be 55mm

But for now I'm using M42 connection. Yes, it causing additional vignetting, but I dont think it should cause any elongation issues, right?

 

What method are you using for polar alignment?

 

I'm using HEQ5 built-in polar scope and iOS application which shows me screen like this - https://prnt.sc/p78o5w - I then rotate on RA axis to match what I see in polar scope with what app shows, and then I move mount up/down left/right with knobs to position polaris according to the app.

I even checking if the polaris still on the circle as indicated by the app like in 2 or 3 hours and it appears to be on the circle in correct position.

But I must add that when I do 2-star alignment, and select first star - it usually not exactly on spot , and sometimes it actually significantly away and not in the field of view, and I had to move mount to find it.

2nd star usually in the field of view.

 

To me, it looks like field rotation.

I would think that in that case the level (severity) of elongation should change depending on exposure length? or maybe I'm wrong?

Edited by axlns, 17 September 2019 - 07:10 AM.

[Stu Todd]

Make sure your primary isn't pinched by the mirror supports.

If they seem not overly tight, it may be cone error?

 

Have a look at https://youtu.be/WatdQlPp22Y for a run down of what may be happening and http://www.sharpcap.co.uk/conesharp to see if this is the case.

Edited by Stu Todd, 17 September 2019 - 07:18 AM.

[AhBok]

You may have multiple issues but I think you need a better polar alignment. You have all the hardware you need. Check out SharpCap (the Pro version is super inexpensive). Then, you can take off PA from the list of problems.

[AhBok]

Oh, and run the Guiding Assistant tool in PHD2. It will give you an idea of how accurate your PA really is.

[OldManSky]

The elongation is consistent across the image (top-right to bottom-left), so I don't think it's collimation.  It's trailing of some kind.

It would help diagnose the issue to know which axis (RA or DEC) the trailing was occurring on.

Can you take one of your images and plate solve it?  That will let you know where North is in relation to the top of your image, and then we can figure out if it's RA trailing (PE or slippage the guider can't correct) or DEC trailing (backlash the guider can't correct or field rotation).

[axlns]

Oh, and run the Guiding Assistant tool in PHD2. It will give you an idea of how accurate your PA really is.

yes, I'm running guiding assistant. I think it shows accuracy within 1 arc-minute

also, I see the following in PHD Log Viewer for my last guiding session - https://prnt.sc/p79r23 -it says PA error is 0.4 arcminute, which I think is pretty small error, right ?

[Alen K]

As stated there are multiple problems. How do you currently polar align?  I ask because your polar axis would have to be significantly misaligned to see trailing in a five second exposure at that focal length. And you have said it didn't matter how long you exposed you see the same arc length in the corners, although you didn't provide evidence of that. (Btw, you can't have eight corners. Four of your named corners are actually sides.)

 

I'm not saying it's not field rotation because there hasn't been enough information presented to rule it out. But it wouldn't be the first time field rotation was blamed on arcs in the corners but was not the actual cause.

 

PS. After I posted I saw your last reply. Unless you have some other problem with your mount (I did not examine your guiding logs) then you were polar aligned accurately enough not to see field rotation in a five second exposure.

Edited by Alen K, 17 September 2019 - 08:12 AM.

[axlns]

As stated there are multiple problems. How do you currently polar align?  I ask because your polar axis would have to be significantly misaligned to see trailing in a five second exposure at that focal length. And you have said it didn't matter how long you exposed you see the same arc length in the corners, although you didn't provide evidence of that. (Btw, you can't have eight corners. Four of your named corners are actually sides.)

 

I'm not saying it's not field rotation because there hasn't been enough information presented to rule it out. But it wouldn't be the first time field rotation was blamed on arcs in the corners but was not the actual cause.

 

PS. After I posted I saw your last reply. Unless you have some other problem with your mount (I did not examine your guiding logs) then your are polar aligned accurately enough not to see field rotation in a five second exposure.

I explained in one of the replies above how I do my polar alignment, in short I use iOS app called PS Align Pro.

Analyzing log of last guiding session in PHD2 Log Viewer says PA error is 0.4' - I believe this indicates good polar alignment.

Regarding saying "top middle corner"  - yeah, my bad I should say "top middle side" - I'll edit my original message

Regarding evidence that elongation are the same, I believe I actually gave that evidence in original message

 

You can compare these screenshots of bottom left corner of the frame

 

5sec exposure - https://prnt.sc/p75y38

60sec exposure - https://prnt.sc/p761e2

 

to me elongation looks about the same

Edited by axlns, 17 September 2019 - 08:27 AM.

[Alen K]

I explained in one of the replies above how I do my polar alignment, in short I use iOS app called PS Align Pro.

Analyzing log of last guiding session in PHD2 Log Viewer says PA error is 0.4' - I believe this indicates good polar alignment.

Regarding saying "top middle corner"  - yeah, my bad I should say "top middle side" - I'll edit my original message

Regarding evidence that elongation are the same, I believe I actually gave that evidence in original message

 

You can compare these screenshots of bottom left corner of the frame

 

5sec exposure - https://prnt.sc/p75y38

60sec exposure - https://prnt.sc/p761e2

 

to me elongation looks about the same

My bad. Missed them both. The conclusion in my PS remains. I used to do One hour exposures (on 35mm film) at 762mm focal length, so I have an appreciation for how much PA error affects field rotation: not as much as everyone thinks.

[axlns]

The elongation is consistent across the image (top-right to bottom-left), so I don't think it's collimation.  It's trailing of some kind.

It would help diagnose the issue to know which axis (RA or DEC) the trailing was occurring on.

Can you take one of your images and plate solve it?  That will let you know where North is in relation to the top of your image, and then we can figure out if it's RA trailing (PE or slippage the guider can't correct) or DEC trailing (backlash the guider can't correct or field rotation).

I learned whats plate solving is, thanks

Here plate solved image (did it in PixInsight) - https://prnt.sc/p7bgdt

[axlns]

My bad. Missed them both. The conclusion in my PS remains. I used to do One hour exposures (on 35mm film) at 762mm focal length, so I have an appreciation for how much PA error affects field rotation: not as much as everyone thinks.

would field rotation be more severe on 5sec vs 60sec frames? I guess it should be right?

but I'm not seeing it, the elongation getting more severe in the corners and it is in different directions

 

but the problem in the thread you mentioned https://www.cloudyni...ut-image-train/ looks very similar to mine. and I dont think they came to conclusion that this is field rotation, or maybe I misread?

[Alen K]

but the problem in the thread you mentioned https://www.cloudyni...ut-image-train/ looks very similar to mine. and I dont think they came to conclusion that this is field rotation, or maybe I misread?

In the end, they did not. But many of the initial prognoses were field rotation as the cause. Whenever you get curving in the corners, it's the quick go-to culprit. But in that case it was not correct. I don't think it is in your case either. But that's just my opinion. Opinions vary.

[axlns]

Make sure your primary isn't pinched by the mirror supports.

If they seem not overly tight, it may be cone error?

 

Have a look at https://youtu.be/WatdQlPp22Y for a run down of what may be happening and http://www.sharpcap.co.uk/conesharp to see if this is the case.

just giving feedback on this advice. I just checked primary mirror and its supports and confirmed that they are reasonable loose. not overly thight for sure as I can rotate them a bit, and they are not tight; so I guess we ruled out another possible reason.

I need to understand what cone error is, thanks for the links provided.

but can cone error have this effect on stars in the corner?

[Galaxyhunter]

would field rotation be more severe on 5sec vs 60sec frames? I guess it should be right?

but I'm not seeing it, the elongation getting more severe in the corners and it is in different directions

 

If it is true field rotation, the longer the exposure, the more trailing will show up.   Is the star that you are guiding on in the FOV of the main image?

[axlns]

If it is true field rotation, the longer the exposure, the more trailing will show up.   Is the star that you are guiding on in the FOV of the main image?

yes, according to my observation that radial trailing in the corners for 5sec and 60sec exposures seems to be the same.

next time I'll have clear sky night I'll try longer exposures, maybe 5 or 10 minutes - just to prove that. never tried this because I was not able to get good results even with 60seconds

 

regarding your other question - good question. I must admit that for this last session I cant be really sure, as I used my finder scope which I did not aligned with my main scope, so guiding star might be located outside field of view of main camera. 

could this it be a reason?

 

ps. previously I used guide scope and aligned with main scope it, but that guiding scope it too heavy and I dont have enough weights to balance RA axis, ordered additional weights to resolve this but it did not arrive yet

 

pps. examples of 5sec and 300sec exposures for some previous session (2 weeks ago or so)

5 sec - https://prnt.sc/p7ebid

300 sec - https://prnt.sc/p7ec2w

again elongation seems to be the same 

and during that previous session I had "triangle" stars, you can see from screenshot. not sure why. is this collimation issue?

Edited by axlns, 17 September 2019 - 12:59 PM.

[jdupton]

axlns,

 

   There are three main problems present in your images from what I can tell. In order of approximate severity, you are seeing the effects of:

• Incorrect spacing from the Coma Corrector to the focal plane. This in introducing Sagittal Astigmatism into your images. That is the main cause of the elongations you see. Only changing the Coma Corrector's spacing will fix this.
   
• Differential Flexure is showing up in the images but is on the borderline of becoming visible in your 60 second images. Your image scale is about 1.22 "/px and your drift rate is approximately 1.14 px/min. Once you fix the Coma Corrector spacing issue, you may start to notice the drift cause by the Differential Flexure. Polar Alignment error looks to be reasonable. The combined effect of DF and PA show that your PA is less than 7.1 arc-minutes. It is probably a good bit less than that as most of the drift error is probably due to the DF.
   
• Your focus is soft at best. You can clearly see the defocused outline of the secondary mirror in your stars. Try to get better focusing and it will then be easier to see the effects of CC back-spacing, DF, and PA errors.

   Even though you may have set the back-focus point of the Coma Corrector to exactly 55.000000 mm, small errors in manufacturing tolerances will force a slightly different back-focus requirement for each system. Upon Plate Solving your first image, I found it gave a focal length of 1006 mm rather than 1000 exactly. Combined with tolerances in the CC lenses, it is common to have to use spacers to adjust the back-focus slightly until your stars meet your criteria.

 

   Here is a PixInsight rendition of the center, sides and corner of the first frame you shared showing the star shapes with the Sagittal Astigmatism aberration and defocus.

 

 

   You can compare the star shapes to the following link which describes the types of optical aberrations that can appear in our images.

https://www.handprint.com/ASTRO/ae4.html

 

   Note that this aberration would be present in even 0.5 second images as it has nothing to do with the star drift in the image. To adjust spacing with all threaded adapters, you may have to resort to using thin spacer rings such as those made by Baader. (https://agenaastro.com/baader-t2-delrin-spacer-ring-set.html)

 

 

John

Edited by jdupton, 17 September 2019 - 02:36 PM.

[Alen K]

regarding your other question - good question. I must admit that for this last session I cant be really sure, as I used my finder scope which I did not aligned with my main scope, so guiding star might be located outside field of view of main camera. 

could this it be a reason?

Field rotation while guiding is always around the guide-star, which is the only thing that is certain not to move (other than by guiding errors). If the guide-star is outside the field of view then field rotation arcs will not be symmetrical around the center of your image. They will be highly asymmetrical. In your image they are more-or-less symmetrical. Their symmetry is yet another indication that the arcs in the corners are not caused by field rotation, since if that were the cause the center of your image near the core of M31 would have been the guiding target, which would be highly unlikely.

[TrustyChords]

axlns,

 

   There are three main problems present in your images from what I can tell. In order of approximate severity, you are seeing the effects of:

• Incorrect spacing from the Coma Corrector to the focal plane. This in introducing Sagittal Astigmatism into your images. That is the main cause of the elongations you see. Only changing the Coma Corrector's spacing will fix this.
   
• Differential Flexure is showing up in the images but is on the borderline of becoming visible in your 60 second images. Your image scale is about 1.22 "/px and your drift rate is approximately 1.14 px/min. Once you fix the Coma Corrector spacing issue, you may start to notice the drift cause by the Differential Flexure. Polar Alignment error looks to be reasonable. The combined effect of DF and PA show that your PA is less than 7.1 arc-minutes. It is probably a good bit less than that as most of the drift error is probably due to the DF.
   
• Your focus is soft at best. You can clearly see the defocused outline of the secondary mirror in your stars. Try to get better focusing and it will then be easier to see the effects of CC back-spacing, DF, and PA errors.

   Even though you may have set the back-focus point of the Coma Corrector to exactly 55.000000 mm, small errors in manufacturing tolerances will force a slightly different back-focus requirement for each system. Upon Plate Solving your first image, I found it gave a focal length of 1006 mm rather than 1000 exactly. Combined with tolerances in the CC lenses, it is common to have to use spacers to adjust the back-focus slightly until your stars meet your criteria.

 

   Here is a PixInsight rendition of the center, sides and corner of the first frame you shared showing the star shapes with the Sagittal Astigmatism aberration and defocus.

 

Image46_mosaic.jpg

 

   You can compare the star shapes to the following link which describes the types of optical aberrations that can appear in our images.

https://www.handprint.com/ASTRO/ae4.html

 

   Note that this aberration would be present in even 0.5 second images as it has nothing to do with the star drift in the image. To adjust spacing with all threaded adapters, you may have to resort to using thin spacer rings such as those made by Baader. (https://agenaastro.com/baader-t2-delrin-spacer-ring-set.html)

 

 

John

Good stuff. I don't want to hi-jack the thread but generally do you reduce Sagittal Astigmatism by decreasing the distance to sensor?

[jdupton]

TrustyChords,

 

Good stuff. I don't want to hi-jack the thread but generally do you reduce Sagittal Astigmatism by decreasing the distance to sensor?

 

   I do not know if there is a general rule or not. I have seen a discussion specifically regarding back-spacing on the EdgeHD scopes that if you are too close in, you get Tangential Astigmatism and if too far out, you get Sagittal Astigmatism. I have verified that that is true on my own EdgeHD scope.

 

   In this case, I would probably try moving closer in by several millimeters (3 to 5) to see if it switched over from Sagittal to Tangential. Then you know exactly which way to adjust and based on the experiment, can try a few different spacings until it meets your own criteria.

 

 

John

[View2]

Flexure ?

[axlns]

   There are three main problems present in your images from what I can tell. In order of approximate severity, you are seeing the effects of:

• Incorrect spacing from the Coma Corrector to the focal plane. This in introducing Sagittal Astigmatism into your images. That is the main cause of the elongations you see. Only changing the Coma Corrector's spacing will fix this.

John, first of all, thank you very much for such detailed and thoughtful answer. Lot of things to consider. Also thanks for the links you shared, they are VERY useful.

 

However, I'm not quite sure how to correct the spacing to make it smaller. At the moment I'm on minimum spacing already.

My optical train is Sony A99, then T-ring (M42), then Baader MPCC III, then 2'' focuser

 

I found that flange focal distance of Sony A99 is 44.5mm (found it here https://wolfcrow.com...for-your-camera), I measured the optical lenght of T-ring I have and it is 10.5mm; so I'm exactly at 55mm; and I cant go lower, maybe only higher with spacing rings I ordered (they did not arrived yet though).

 

Also, after closer review of the images from my one of my previous sessions, I found that I dont see that Sagittal Astigmatism (I think I dont see it, but maybe i'm wrong), and my setup did not changed, except that last time I used finder scope not aligned perfectly with main scope

 

Please check these images - I believe they do not have that "radial" Sagittal Astigmatism. I think maybe there are some "trends" to it, but it is not even near to what I see on my last images. There is still really bad elongation, but maybe it will give you a clue that this is not the coma corrector spacing after all?

This is 300seconds exposure - https://prnt.sc/p7jtyp

This is 2seconds exposure - https://prnt.sc/p7jub9

Here you can download (51mbytes) zip file with 3 images in FIT format -300seconds, 2seconds and 0.5seconds - all 3 I believe shows similar pattern for elongation, but it does not look like that radial pattern from my last session - https://www.sugarsyn...09685078_675604

 

Also another question - could these issues be caused by MPCC not exactly aligned within focuser? It is not threaded connection to focuser (my focuser does not have this option), so I just use 2'' connection in focuser, and perhaps under the weight of camera it had some tilt? Could this produce such issues?

 

 

Even though you may have set the back-focus point of the Coma Corrector to exactly 55.000000 mm, small errors in manufacturing tolerances will force a slightly different back-focus requirement for each system. Upon Plate Solving your first image, I found it gave a focal length of 1006 mm rather than 1000 exactly. Combined with tolerances in the CC lenses, it is common to have to use spacers to adjust the back-focus slightly until your stars meet your criteria.

Thats interesting, but can it be not 1000mm precisely just because for example I set the primary mirror slightly more distant from the secondary when I did collimation? Will that produce these type of astigmatism ? and how that difference affect back focus length?

 

Your focus is soft at best. You can clearly see the defocused outline of the secondary mirror in your stars. Try to get better focusing and it will then be easier to see the effects of CC back-spacing, DF, and PA errors.   

Yes, focus was wrong in the beginning of session, but I get it corrected in the middle - for example if you try another image, say L_M31_00163.ARW from the archive you downloaded - you should not see there focus issues, can you please check?

 

• Differential Flexure is showing up in the images but is on the borderline of becoming visible in your 60 second images. Your image scale is about 1.22 "/px and your drift rate is approximately 1.14 px/min. Once you fix the Coma Corrector spacing issue, you may start to notice the drift cause by the Differential Flexure. Polar Alignment error looks to be reasonable. The combined effect of DF and PA show that your PA is less than 7.1 arc-minutes. It is probably a good bit less than that as most of the drift error is probably due to the DF.

I need to learn what differential flexure is. At the moment I dont understand it, I'll do my research, perhaps you can suggest some links? Is that occur when field of view of guiding camera and field of view of imaging camera are offset by somewhat large distance? Thanks!

Edited by axlns, 18 September 2019 - 12:33 AM.