EdgeHD 11 Imaging Blurry at F/7 and F/10? - Forums

Hi everyone,

I have been imaging with my new EdgeHD 11 over the last couple of months predominantly though a Hyperstar at F/2 along with a ZWO ASI2600MC PRO. The images I have been able to capture at F/2 are stunningly sharp. The 540mm focal length and 3.76 um pixel size gives approximately 1.43 arc seconds / pixel, which is right in the middle of the ideal Nyquist Limit range under average seeing conditions (Bortle 5 skies here).

Over the last few days, I managed to acquire the parts to image from the back at F/7 and F/10. I am reaching a back focus of 146.05mm with my imaging train (using a Celestron OAG and the appropriate spacers that come with the camera). I also use a Celestron focus motor, along with NINA’s autofocus utility (both NINA vanilla and Hocus Focus) and in each case I get a reasonably good hyperbolic curve. I’ve also double checked this with “manually” adjusting the focus motor to a similar point. I’ve also tested binning at 1x1, 2x2 and even 3x3 (particularly that last one brings it much closer to 1.2 arc seconds / pixel at F/7).

Here’s the crux of the issue: in all cases the images appear to be very blurry at F/7 and worse at F/10. I’ve tried imaging several reasonably easy targets (NGC 7000, M 42, NGC 7635). I would love feedback from others here who have had experience imaging at F/7 or F/10 (even better if it is with a similar-ish rig).

For example, here’s NGC 7635 @ F7 @ 2x2 binning (30 x 240” subs):
https://www.astrobin.com/49y01t/

And here’s the same version sharpened (using an ML based sharpener), which is what I would expect the original to be closer to:
https://www.astrobin.com/0fjc4e/

And here’s NGC 7635 @ F10 @ 2x2 binning (45 x 240” subs), which just looks terrible (and yet I was at the sharpest focus I could get both manually and through NINA / Hocus Focus):
https://www.astrobin.com/rjsp3r/

Even this section of NGC 7000 @ F10 @ 2x2 binning (45 x 120” subs) didn’t turn out great:
https://www.astrobin.com/ii59dw/

Particularly compared to what I see at F2:
https://www.astrobin.com/hq1bed/

M 42 @ F7 1x1 binning (10 x 120” subs) also looks rather blurry:
https://www.astrobin.com/5ernz6/

So I would be very grateful if someone can help me understand if my expectations are off and this is the level of blurriness expected with my EdgeHD 11 and camera combination at F/7 and F/10, or whether I do have a focus problem with my telescope / imaging train / camera.

Thanks so much!

On your hyperstar picture it seems to be a bit out of focus in addition to tilt and backfocus issues.

As for the slower focal ratios, are you confirming focus with a bahtinov mask when doing it manually? In addition it would help to know what your imaging train is for the slower focal ratios.

Thanks so much for the prompt reply. That's interesting that you were able to identify tilt and back focus issues with the Hyperstar. I am using the Hyperstar, the Starizona modular filter slider, and then the camera, so everything basically off the shelf. Any tips on what I can do dig in would be very helpful.

As for the slower imaging trains, here's what the setup looks like:
(Reducer - in case of F/7)
Baffle Nut - 12.7mm (I believe)
SCT Adapter - 25.3mm
OAG - 29mm
42mm Male Adapter - 12.5mm
Medium T Spacer - 11.55mm
ZWO 16.5mm Spacer - 16.5mm
ZWO 21mm Spacer - 21mm
ZWO ASI2600MC PRO - 17.5mm

Are you noticing tilt issues in the slower f-ratio images as well?

As for how I am verifying focus manually, it's by looking at the Star HFR in NINA and getting it down to the lowest value.

I use both the ASI2600 and 6200 cameras on my C11”EdgeHD @f/10 and at f/7 in the past. I also use NINA with the HF for auto focusing. Your images are way out of focus but I don’t see a bad tilt problem. However, forget using the NINA HFR value for now. As suggested already, use a Bahtinov mask and manually focus. Also, recheck your backfocus. This video might help….https://m.youtube.com/watch?v=i1uccN3BxMg.

Make sure the primary mirror locks are loose and that the Celestron autofocus motor is not binding. This is a know issue and mind it. I got rid of it for a better device.

If your backfocus is correct and the autofocuser motor can move freely, a Bahtinov mask should do the trick and give you a very sharp image. Once you get that, run NINA with HF to generate a nice focus curve with a R2 value around .97 or higher, with 1 being perfect.

The very short focal length afforded by the Hyperstar allows a lot of slack in regard of seeing and optical quality overall (read collimation and tube currents). Once you're on native focal lengths you're would need to go way down in sampling to, I would say, at least 1"/px native (i.e., as sampled at focus) to hope to achieve similar quality of focus in less the optimal seeing conditions. I see you're way oversampling at between 0.4 to 0.5"/px and my hunch is you need to go down, way down in your sampling to hope to achieve something close to the sharpness you observe at f/2. Ideally you would need an imager with 9 um pixels but these are rarer than hen's teeth nowadays.

Just for the heack of it, what is the fwhm reading you get from the images at f/7 and f/10?

Hi Bruce, I have found that your backfocus is undercorrected by 12.7mm. Celestron's specifications for the Edge HD Series state the backfocus on the Edge HDs in F/7 & F/10 is not from the end of the baffle nut, but the end of reducer in F/7 and at the end of the threads in F/10. You are most likely getting very bad, "blurry" results as a result of being undercorrected so badly.

The Edge HD 11 has a perfect backfocus of 146.05mm at both focal lengths, but because of your miscalculation adding the 12.7mm from the baffle nut on to your image train, you are effectively at 133.35mm, which at 12.7mm undercorrection gives a star spot diameter error of nearly 13 times the normal amount!

I hope this possibly solves your issue, please come back after adjusting and report your findings I'd love to see if it worked!

Michael.

Hi Bruce, I have found that your backfocus is undercorrected by 12.7mm. Celestron's specifications for the Edge HD Series state the backfocus on the Edge HDs in F/7 & F/10 is not from the end of the baffle nut, but the end of reducer in F/7 and at the end of the threads in F/10. You are most likely getting very bad, "blurry" results as a result of being undercorrected so badly.

The Edge HD 11 has a perfect backfocus of 146.05mm at both focal lengths, but because of your miscalculation adding the 12.7mm from the baffle nut on to your image train, you are effectively at 133.35mm, which at 12.7mm undercorrection gives a star spot diameter error of nearly 13 times the normal amount!

I hope this possibly solves your issue, please come back after adjusting and report your findings I'd love to see if it worked!

Michael.

Hi Michael

My backfocus is just fine. I assume you were addressing ashastry who made the original post.

Bruce

My first images with an SCT at f/7 and f/10 had similar, blob-shaped stars back in 2012-2013. The most significant culprit turned out to be autoguiding accuracy. Even if the telescope is perfectly in focus, wobbling of the guide star will enlarge star shapes and make a blurry mess out of the image.

I ended up using an ONAG device from Innovations Foresight for many years, and with great success. I now also own a Celestron OAG. Both devices work just fine with PhD2 guiding. May I ask what are typical RMS error values (either in pixels or arcseconds) you get with your EdgeHD and the OAG?

Bruce Donzanti:
Hi Bruce, I have found that your backfocus is undercorrected by 12.7mm. Celestron's specifications for the Edge HD Series state the backfocus on the Edge HDs in F/7 & F/10 is not from the end of the baffle nut, but the end of reducer in F/7 and at the end of the threads in F/10. You are most likely getting very bad, "blurry" results as a result of being undercorrected so badly.

The Edge HD 11 has a perfect backfocus of 146.05mm at both focal lengths, but because of your miscalculation adding the 12.7mm from the baffle nut on to your image train, you are effectively at 133.35mm, which at 12.7mm undercorrection gives a star spot diameter error of nearly 13 times the normal amount!

I hope this possibly solves your issue, please come back after adjusting and report your findings I'd love to see if it worked!

Michael.

Hi Michael

My backfocus is just fine. I assume you were addressing ashastry who made the original post.

Bruce

My bad, I was adressing ashastry, I mistook your gold nameplate for being an OP plate lol

Michael.

My first images with an SCT at f/7 and f/10 had similar, blob-shaped stars back in 2012-2013. The most significant culprit turned out to be autoguiding accuracy. Even if the telescope is perfectly in focus, wobbling of the guide star will enlarge star shapes and make a blurry mess out of the image.

I ended up using an ONAG device from Innovations Foresight for many years, and with great success. I now also own a Celestron OAG. Both devices work just fine with PhD2 guiding. May I ask what type of autoguiding you use? What are typical RMS error values (either in pixels or arcseconds)?

Guiding errors are non-circular and would appear as lines and "shake" in the stars, and the smearing of details, his stars and details are stationary in his photos, this appears to be an optical issue.

Really poor guiding can still produce round, blurry stars if corrections/spikes are severe in both RA and Dec. I'd still like to see his RMS error values to confirm it isn't autoguiding or exceptionally poor seeing (which could also be a culprit).

@SemiPro @Bruce Donzanti Thank you for your feedback. I agree, it seems like I should go back to trying to focus with a Bahtinov mask and see if that makes a difference.

@andrea tasselli I agree that from a theoretical perspective, that ~1 arcseconds / pixels - ish range seems ideal, but @Bruce Donzanti actually has several shots on astrobin with a very similar setup imaging with the ASI2600MC PRO binning at 1x1. It's seems there is something else amiss here.

@Michael Legary I probably didn't articulate the back focus distances well there; I am essentially following what's highlighted in the EdgeHD white paper and the OAG manual: apologies, what I called the baffle nut in the post above is the adapter plate. Sorry for the confusion on that. In both F/7 and F/10 the adapter plate goes on top of the threads (the reducer in case of F/7, or the baffle lock nut in case of F/10). I've measured the image train length, and it's hitting 146.05mm.

@rhedden Thank you for your feedback. Definitely something for me to consider, I am getting RMS values around ~0.6-0.8" for RA and ~0.5-0.7" for DEC with a total around ~0.8-1", which is probably higher than ideal.

Some ideas:

Are all your subs identical? Could it be that you focus well and then your focus slips over the course of imaging? This would show if your first sub is better than the rest. Please upload your nina focus curve showing the HFR, your nina auto focus settings, and the image of the best focus during the focus routine.

In Nina autofocus, select to focus only in the middle of the frame, this way your focus will not be impacted by blurry corner stars. (set inner crop ratio to 0.7 or 0.8)

Do you overshoot in Nina autofocus routine? In my C8 it is absolutely necessary to overshoot. (backlash compensation method = overshoot, and put a number in the backlash Out field, say 10 or 20 times your autofocus step size)

Check whether you focus at the same focus point with nina and with bahtinov mask.(if you don't have one, you can make a simple one with cardboard and a sharpie)

Use the nina abberation inspector from hocuspocus and run it.

How does your collimation look? Barlow, bright star, centre field of view, very short exposures (guide cam is ideal for this) and defocus slightly just to make out that the star is turning intoa ring/donut - is the ring in good shape or is the hole off-centre?

RMS guiding error of 0.8"-1.0" is on the high side, and it is enough to blur the stars slightly with an 11" EdgeHD - but it can't account for the amount of blur you're seeing. I agree with the others that there is possibly an autofocus issue here.

The suggestions made by Cosmetatos are good ideas. If it turns out the autofocuser is not finding the proper position even after adjusting backlash, you might also want to try giving the autofocuser a little more more time between focus steps, especially when the weather is cold.

When my SCT is out of focus I see some darkening in the center of the stars as a result of secondary obstruction (doughnut effect), I don't see this on your images...puzzling. Here is an out of focus M42 with C14 and a reducer, if you zoom in you will see the doughnut effect on the stars.

I've also imaged with EdgeHD 11" at f/2, f/7 and f/10. I abandoned Hyperstar but that was a personal choice. I have the Celestron Focus Motor, but also invested in a Moonlite CHL EdgeHD because getting fine focus with the primary mirror focusing mechanism is next to impossible at f/7 and f/10. Yes, it costs a lot, but it sure did solve the problem. Over the past 2.5 years, I have come up with the following workflow which has corrected most of my issues:

1. Set backfocus to 146.05mm and point scope towards zenith. If using monochrome, remember to add distance for your filter thickness.

2. Adjust primary mirror focus with Celestron Focus Motor. Once FWHM (or HFR) at its best, lock the mirror locks.

3. Run autofocus (Hocus Focus) in NINA using the Moonlite focuser.

4. Run Metaguide to ensure collimation is spot-on while *in-focus* (NOT the donut method). Re-run autofocus as needed. It is best to do this without the focal reducer in place.

5. Ensure that collimation with Metaguide remains stable WITH and WITHOUT camera's cooler fan running. Fan vibration is like the tuberculosis of astrophotography: it can mimic anything!

6. Check sensor tilt and adjust as necessary.

I have done these things in that particular order because issues with the earlier steps will throw off the process in later steps (i.e. if your primary mirror is in the wrong position, that will affect field curvature which may throw off tilt calculations, etc). Check collimation and fan vibration before adjusting tilt because the former can artificially affect your tilt.

Hope this helps!
Ben

What I can conclude from looking at your images, I do see a small amount of tilt but nothing really important compared to out-of-focus blurr. It is circular and dos not show any tendency towards either RA or Dec direction - at least nothing I can see on the resolution reduced images.
I would check the subs and look for consistency in that blurr-shape - is it always that circular or is that final shape the result of a combination of blurr into various directions - if so, that means it is a guiding or stability problem (I do not think so).
I presume that you subs are just out of perfect focus - despite using the AF system of NINA. I have often a similar situation on my Newton - the NINA AF as well as the Hocus Focus give me a certain result which is not bad but not perfectly the focus optimum. I then correct it manually according to experience in the one or other direction. Like that I often get better HFR data - usually (as example) from AF 2.8 to manally corrected 2.5.
One more source of problems can be the fan which cools the inside - I have one on my RASA11. It works very fine but I need to turn it off before I start critical imaging - the small vibrations result is slightly blurred subs…maybe something to check as well.
The AF system ( whatever you use - ZWO EAF or SestoSenso…) can be another source for inaccurate focusing - I use the SestoSenso as well as the ZWO EAF. The ZWO is less accurate compared to the SestoSenso in my opinion. The focusing mechanism itself can be a problem as well - on my RASA I have the Moonlite which is very fine and precise - before that I had the original Celestron focuser, which was neither precise enough nor repeatable enough. So get the best possible focusing system for critical imaging.
Another observation I made is that even with perfect AF results, the HFR outcome changes - most of the time to worse values, but also sometimes to smaller values. Especially if your target rises in the skies, the diffraction changes and the quality of the subs therefore change as well. That means, I start with HRF of 3.5 at the beginning of and evening with the target at, say, 25 degrees over the horizon and later I get down to 2.5 when the target is about 45 degrees over the horizon. If you have set your focus with the AF routine when the target was at 25 degr - and you do not refocus some time later, you end up with subs which are all slightly out of focus.

These are just my thoughts….
Hope you get to super results soon!
Georg

@Cosmetatos, @rhedden, @drblevy13, @Georg N. Nyman Thank you all so much for your feedback, I really appreciate it. Lots of good input on this thread for me to sift through and apply. I tried a few things last night, including leaving my scope out for much longer, giving some extra breaks between NINA's autofocus steps, and simplifying my imaging train by getting rid of the OAG and instead using a minimal train with the reducer, the celestron T-adapter and the camera (with an external guider).

I have a Bahtinov Mask on order, but unfortunately that won't get here till early next week, and I am looking at cloudy skies here in the PNW for the next two weeks. When we get back to clear skies here a some point I will get to hopefully testing a few more things out, including confirming the focus numbers across NINA and the Bahtinov Mask.

I also suspect that my Celestron Focus Motor just isn't being precise enough as @Georg N. Nyman suggested, particularly with the super heavy 11" mirror, so I will invest in a better focuser and see if that makes a difference.

@Marc Dickinson Thanks for sharing that image through your C14. I would love to see what the in-focus image looks like so I have a good reference.

In the meantime, here's how far I got with all my experiments, and it's definitely progress, so again thank you all for your valuable feedback!
https://www.astrobin.com/2fe9zb/

You're making great progress! I would do exactly what you are doing and simplify. I would then probably do some simple images of the Moon, maybe crater Plato, and gain confidence in the focus and resolution.

Poor seeing, poor guiding, poor focus, thermals inside the tube.

you have to tackle these first, there’s just no way around it besides the poor seeing.

at 2800 or 1960mm guiding should be under .5” even closer to .25”

with an SCT you might have to re-focus every 30 minutes!!! You cannot guarantee good focus across the field by looking at HFR between subs if the seeing is bad. That number will be jumping around.

I second seeing. Not sure where you live @ashastry, but where I'm at in the northeastern US, it's not uncommon to get NINA AF HFR measurements double what I call 'good' and triple my best. Sometimes almost that much of a difference through a single night. As @Georg N. Nyman described, moving towards or away from the horizon will also have an effect, and poor seeing results in poor guiding, so a bit of vicious-cycle to boot.....

Cheers,
Scott

Circling back on this. After trying a bazillion things, including Bahtinov mask based focusing, more accurate guiding, trying a different camera, etc. I continued to capture images that I felt were just not crisp enough. No matter what configuration I tried (OAG, no-OAG, and appropriate adapters respectively) at 146.05mm, and binning at 1x1, 2x2 or 3x3, I would only get images like on the right.

So I finally invested in a micro-focuser (the Pegasus Prodigy) and I think it has revealed the source of the problem: it appears to be a back-focus issue. With the micro-focuser in place (reaching the recommended 146.05mm back focus), and the primary mirror locked at the same place I was achieving best focus previously, I was able to achieve a much sharper image (on the left) than I had been able to capture before, except that to get there I had to reach a distance of 154.17mm by drawing out the focus tube, quite a bit further away from the 146.05mm.

I don't know where the source of the issue is, but I have a thread going with Celestron.

Circling back on this. After trying a bazillion things, including Bahtinov mask based focusing, more accurate guiding, trying a different camera, etc. I continued to capture images that I felt were just not crisp enough. No matter what configuration I tried (OAG, no-OAG, and appropriate adapters respectively) at 146.05mm, and binning at 1x1, 2x2 or 3x3, I would only get images like on the right.

So I finally invested in a micro-focuser (the Pegasus Prodigy) and I think it has revealed the source of the problem: it appears to be a back-focus issue. With the micro-focuser in place (reaching the recommended 146.05mm back focus), and the primary mirror locked at the same place I was achieving best focus previously, I was able to achieve a much sharper image (on the left) than I had been able to capture before, except that to get there I had to reach a distance of 154.17mm by drawing out the focus tube, quite a bit further away from the 146.05mm.

I don't know where the source of the issue is, but I have a thread going with Celestron.

Backfocus is measured from the end of the nut lock, ignoring the nut lock's backfocus, not the end of the baffle tube like I suspected. I think you can make it even sharper with more tuning.

Additionally, it is measured from the end of the reducer when you have the reducer equipped.

@Michael Legary I am achieving focus at 154.17mm from the end of the 0.7x Reducer (at the end of the threads), not 146.05 which it should be.