C 9.25 EdgeHD / smaler stars oblong / larger stars mostly ok

Jacob,

I agree with everything Ben has said above. I wasted countless hours trying to work out what was up with my Edge HD 8 and the 0.7 x reducer. I tried OAG, a myriad of guide scopes to rule out guiding and then focused in on tilt or collimation.

I found it almost impossible to tweak the tilt using the ZWO tilt plate and CCD inspector and again waisted countless hours of my very scarce clear skies (note there are better off brand tilt plates for the camera but I didn't take the plunge). Then one evening I ditched the focal reducer and started shooting at f10 and everything worked fine. So for now I'm going to shoot at f10 and BIN 2x2. I use and ASI 2600 MM Pro so the field is still pretty wide and as I only use the Edge for small targets that's fine for me. I do plan to give the reducer another go and I don't want to blame the oblong stars on the optics it's most likely just introducing another tilt component into the system and at 1400 mm focal length the image scale is not forgiving! Also compared to my William Optics FLT 132 the image train on the HD 8 is much longer and far less robust so all the more opportunity for tilt/flexure.

The reducer wasn't a total waste I still managed some pictures of M57 and M27 but I had to correct the stars using an offset filter which is far from ideal.

Let us know how it works out.

Thanks

John

P.S my stars looked exactly the same as your - if I recall I needed a 2 by 1 or 3 by 2 (faded) pixel offset filter to correct them

Thanks Ben I’ve always been a fan of Gerd I think I bought some of his first filters way back in the early 2000s. Perhaps I’ll donate some more for the tilt plate!!!

Hi Jacob,

I have a similar setup but I have an EdgeHD 11", Moonlite Edge CHL focuser and ASI 2600MC Pro.  When I look at your images, even the larger stars are oval; it is not limited to the smaller ones.  It is just easier for you to see on the smaller ones.

If your backspacing is correct, most likely you are dealing with sensor tilt and potentially inaccurate collimation.  The reason why I think it is mostly sensor tilt is, firstly, that ZWO has had major issues with this, and secondly, all your stars are oblong in a similar direction.  Collimation may not be perfect because even the middle of your center image is affected, and typically with sensor tilt,  stars dead center of the sensor are spared from tilt.

What process did you use for collimation?  The defocused star method (donut) is not ideal but yet highly publicized.  I would strongly suggest using the free program Metaguide, which has a bit of a learning curve, but is well worth it.  Check out this thread:  https://www.cloudynights.com/topic/659381-need-help-collimating-edge-hd/

This YouTube video from the creator of Metaguide also is phenomenal (he is also PhD in optics, so reliable source):  
https://www.youtube.com/watch?v=hZV43ZHNUKM&t=1475s

Once you have gotten collimation as good as you can get it there, then comes the issue of sensor tilt, which I personally found much harder to learn conceptually and is a known problem with ZWO cameras across the board.  Check out this thread for details:

https://www.cloudynights.com/topic/775953-asi6200-tilt-adjustment/

The program used in that thread to detect sensor tilt is called CCD Inspector.  It has a 30 day trial period, but after that is quite expensive.  Similar methods can be done using PixInsight and ASTAP, the latter of which is free.  Major pitfall though is ASTAP is significantly slower to compute the tilt than CCD Inspector so the process may take much longer.

I went so far as to build the laser jig referenced in that thread, but in hindsight, iterative star testing was still the best way to work it out.  Be prepared to sacrifice a couple clear nights to work out the technicalities.  This can all be a bit overwhelming for sure.  I keep trying to work out the details on my own setup too.  The pursuit of perfection is never-ending!  If you need more help, I can try to assist.

Clear skies,
Ben

Hi Ben,
Hi all,

thank you for the support. Especially two observations are quite vexing for me:

- a few days ago a did a fresh collimation with the large donat method as good as i can manage. Before that, collimation was slightly off. Interestingly pictures were better before that, because oblong stars are only detected more or less are detectalbe in two corners instead overall.

- prominent oval stars seem to differ slightly from image to image. Mostly in upper right corner it is very bad. Sometimes more or less equally distributed at upper border, sometimes lower picture area.

To make matters worse, some condensation has been cultivated on the interior side of corrector class. Hence i am doing this only for 1,5 years all this equipment is fairly new and i am not that experienced with it. I deducted following procedure:

- careful cleaning job
- new collimation with Metaguide
- image with 0° rotation (via rotator)
- image with 180° rotation
- compare with ccd inspector (if more or less identical equals sensor tilt, of not it must be something else)
- if so, remove FR and start over ...

I recon (after some reading) sensor tilt might be curable. What would you recommend if FR is the problem like others reported. Using our gear at f10 instead of f7 word double aquisition times. I am not sure this is acceptable because we live in an area with sparsly clear skies...

Thanks and i will report.

Cheers Jacob

I, like many others, have struggled to get good star profiles with the SCT, especially with the larger ones like the C11" EdgeHD, which I have, and the C14" EdgeHD.  I cannot disagree with much that has been suggested but from my own experience and those from others who also helped me, I would ditch the Celestron focal reducer.  Maybe it is not as much an issue on the C9.25 but it is essentially frustrating to use it on the C11 and C14 unless you are willing to do significant cropping.  And just because you have carefully measured to get the backfocus correct per Celestron specs, does not mean it is accurate for your setup.   You may still need to tweak it by fractions of a mm to get it spot on, if you want to.  I am under Bortle 7 skies and only do f/10 imaging even when I only have average or below average seeing conditions and it is fine.  Acquisition time does not have to be double vs. that when imaging at f/7 to get a decent image and I also have minimal clear nights per year.  So, once you have good polar alignment, good guiding, and the best collimation you can do (I only use the eyeball test on a decent star with the camera- no softwire), you will get nice round stars at f/10.  To improve the image at f/10, I dither each frame and sometimes do deconvolution, but only on those nights where I am oversampling due to below average seeing.   I realize camera tilt is a whole other nightmare to deal with as I have had that problem as well, but you don't need perfection.    Currently, the tilt on my ASI6200mm on my C11" EdgeHD varies from 1-3%, which I can live with as you hardly can notice it.  And the little bit you can see essentially goes away as you have to crop a tiny bit due to the dithering effect of the stacked frames.

Best of luck!  This can all be frustrating but you will reach a happy point as we all eventually do.

Bruce

Thanks, John.

I tried that a few times, but frankly, I could not tell much of a difference with either camera.  So, maybe it is just my not-so-perfect eyesight.

Jacob,

In my case I know adding the f7 reducer was the cause of the problem. I just switched back to my Edge HD8 for Galaxy Season and ran it at f10 no issues perfectly round stars. Oblong stars at f7, but I don't think its the reducer I think its the image train flexure or tilt. Though why I'd get more flexure with 105mm than 133mm I don't know but I do! Anyway I just ordered an more sturdy spacer so let's see how that goes I will report back as soon as I have the results!

Give f10 a try and BIN 2x2 even from my Bortle 7 backyard there was a lot of data on M1 so I'm expecting decent images on brighter targets!

John

Hi Ben and John,

thank you very much! What do you think; does this distortion come with every Celestron FR? I am just wonderung …. 

Cheers Jacob

I hope not but if it is it doesn’t appear to be systemic there a plenty of images here taken at f7. Much more likely to be set up, it’s possible that at f7 the system is just more sensitive to flexure/tilt.

Hi all,

this is a head's up on our oval star issues. Please find attached a short documentation (details. ccd-inspecter plots). I am not sure what to do now. F7 and F10 are not ideal. Next step will be another collimation attempt. 

Your insight would very much appreciated. Thank you!

Cheers Jacob

2022-02-19_whatnow.pdf

Hi Jacob,

Looks like you have put A LOT of effort into this.  Very good work though.  A couple of thoughts.

1. It looks like you are trying to run CCD Inspector on images of an open cluster.  While not as bad as a globular cluster, that may still throw off the algorithm CCD Inspector uses to make its calculations.  When trying to get a good read on tilt and field curvature, it is usually best to use a more uniform star field.  Finding a rich star field anywhere in the galactic equator is usually a good starting point.

2. If we assume that the open cluster did not throw off your analysis, then it looks like the most major problem you have across your samples is a large degree of field curvature.  Most common reason for this is inaccurate back focus distance.  What did you use as your starting point for measuring your back focus?  There are tons of posts on Cloudy Nights over this issue with the EdgeHD series.  Starting point needs to be the flat portion of the end of the threads (rear cell for f/10 or end of threads of focal reducer at f/7).

I personally found that the only way to get an accurate read on back focus was to measure the distance from *inside* the focuser baffle tube rather than use an external part of the scope as a marking point.  In the attached picture, I measured from the flat outer portion of the threads of the focal reducer to the most outside portion of my focuser.  Once I knew that distance (in my case 64mm), I subtracted that distance from 146.05 and got 82.05.  I added one more millimeter because of my filter thickness, and then used 83mm as the final distance from the rear of my focuser to the camera sensor.  That substantially improved field curvature.

3.  Once you have your back focus spot on, point the scope to the zenith.  First focus the primary mirror to get the lowest FWHM/HFR you can WITHOUT changing your backfocus distance (in other words, do not use the Nitecrawler for this part).  Make sure your last mirror focus adjustment is turning the focus knob counterclockwise.  This pushes the mirror away from the rear of the scope and against gravity.  This will minimize issues with mirror flop.  Once you have that focused the best you can using the primary mirror, LOCK THE  MIRROR.  You can then do your final focus using the Nitecrawler.

4. Consider redoing your collimation once you have completed #2 and #3 above.

Hope this helps!

Ben

Hi Jacob,

Looks like you have put A LOT of effort into this.  Very good work though.  A couple of thoughts.

1. It looks like you are trying to run CCD Inspector on images of an open cluster.  While not as bad as a globular cluster, that may still throw off the algorithm CCD Inspector uses to make its calculations.  When trying to get a good read on tilt and field curvature, it is usually best to use a more uniform star field.  Finding a rich star field anywhere in the galactic equator is usually a good starting point.

2. If we assume that the open cluster did not throw off your analysis, then it looks like the most major problem you have across your samples is a large degree of field curvature.  Most common reason for this is inaccurate back focus distance.  What did you use as your starting point for measuring your back focus?  There are tons of posts on Cloudy Nights over this issue with the EdgeHD series.  Starting point needs to be the flat portion of the end of the threads (rear cell for f/10 or end of threads of focal reducer at f/7).

I personally found that the only way to get an accurate read on back focus was to measure the distance from *inside* the focuser baffle tube rather than use an external part of the scope as a marking point.  In the attached picture, I measured from the flat outer portion of the threads of the focal reducer to the most outside portion of my focuser.  Once I knew that distance (in my case 64mm), I subtracted that distance from 146.05 and got 82.05.  I added one more millimeter because of my filter thickness, and then used 83mm as the final distance from the rear of my focuser to the camera sensor.  That substantially improved field curvature.

3.  Once you have your back focus spot on, point the scope to the zenith.  First focus the primary mirror to get the lowest FWHM/HFR you can WITHOUT changing your backfocus distance (in other words, do not use the Nitecrawler for this part).  Make sure your last mirror focus adjustment is turning the focus knob counterclockwise.  This pushes the mirror away from the rear of the scope and against gravity.  This will minimize issues with mirror flop.  Once you have that focused the best you can using the primary mirror, LOCK THE  MIRROR.  You can then do your final focus using the Nitecrawler.

4. Consider redoing your collimation once you have completed #2 and #3 above.

Hope this helps!

Ben

following up on this;

a short read caused more confusion (as most times). According to 

https://s3.amazonaws.com/celestron-site-support-files/support_files/edgehd_whitepaper_final.pdf

backfocus distance should be placed within 0.5mm. But this

https://www.innovationsforesight.com/support/celestron-edgehd-back-focus-tolerance/

stated, an offset 20 mm with average seeing should acceptable. Reason I am asking is, adjusting or better put mitigating backfocus distance from 151 mm to 146 mm would cause som trouble, for instance purchase of an thinner oag …

thank you again
cheers Jacob