Improving HFR values [Deep Sky] Acquisition techniques · Jean-David Gadina · ... · 10 · 320 · 1

macmade 3.01
...
· 
·  Share link
I recently started using NINA to take my images.
I'm still discovering and learning a lot, and I began to pay attention to my HFR values.
However, I wonder if the values I'm observing are within the acceptable range.

My current setup is:
  • Celestron NexStar Evolution 6 (1500mm), on a wedge
  • ZWO ASI585MC Pro (2.9um)

I have some over-sampling, and I also read that longer focal lengths typically increase HFR values.

Currently, I'm imaging the Crescent Nebula with a 0.63x reducer. My focal length is 945mm.
My average HFR values are around 4.2px with a 3-minute exposure time.

Is it acceptable for my particular setup, or should I aim at a lower HFR?
If so, what can I improve?

Should I improve guiding, take shorter exposures, or use binning?
Thanks a lot for your advice!

CS

NINA-Crescent-2.png
Edited ...
Like
jhayes_tucson 26.84
...
· 
·  5 likes
·  Share link
The diffraction limited Airy diameter for your F/6.3 system is about 8 microns (7.68 um at 0.5 microns).  If you are measuring a 4.2 pixel half-flux radius with a 2.9 micron sensor, that’s a star diameter of roughly 24 microns, which is 5.3 arc-seconds in the sky.  That’s pretty big and I suspect that unless your seeing is really bad, that you should be able to do better.

There are a lot of factors that go into determining the star size in your integrated images.  First, you need high quality optics and they need to be very well aligned.  Second, you need perfect focus.  Your autofocus curve looks good so that shouldn’t be a problem. Third, you need excellent tracking and really good mechanical stability.  For your scope, if your total rms tracking errors are more than roughly 0.4”, under good conditions, that will be the limiting factor in determining your star size.  Finally, you need a very stable atmosphere.  The good news for you is that with a 6” scope, you should able to tolerate pretty mediocre conditions (around 2”) and still get pretty good results.

Do an experiment and take some very short exposures—less than say 10s.   Use bright stars if you have to and then measure the HFR.  That will tell you how well you autofocus and optical system are working independent of tracking errors.  Also check to make sure that you have the telescope set up on a very stable foundation with no vibration.   Balconies and rooftops are typically terrible—for both vibration and turbulence!

I run a 5” F/6.5 system in Chile under mostly very good seeing and I often get FWHM stars in the range of 1.4” - 1.6” when the seeing is good.  My feeling is that if you tune it up, you should be able to get your system FWHM down to at around 2” ( HFR < 2 px)  with good seeing.

John
Like
aabosarah 9.31
...
· 
·  Share link
John Hayes:
The diffraction limited Airy diameter for your F/6.3 system is about 8 microns (7.68 um at 0.5 microns).  If you are measuring a 4.2 pixel half-flux radius with a 2.9 micron sensor, that’s a star diameter of roughly 24 microns, which is 5.3 arc-seconds in the sky.  That’s pretty big and I suspect that unless your seeing is really bad, that you should be able to do better.

There are a lot of factors that go into determining the star size in your integrated images.  First, you need high quality optics and they need to be very well aligned.  Second, you need perfect focus.  Your autofocus curve looks good so that shouldn’t be a problem. Third, you need excellent tracking and really good mechanical stability.  For your scope, if your total rms tracking errors are more than roughly 0.4”, under good conditions, that will be the limiting factor in determining your star size.  Finally, you need a very stable atmosphere.  The good news for you is that with a 6” scope, you should able to tolerate pretty mediocre conditions (around 2”) and still get pretty good results.

Do an experiment and take some very short exposures—less than say 10s.   Use bright stars if you have to and then measure the HFR.  That will tell you how well you autofocus and optical system are working independent of tracking errors.  Also check to make sure that you have the telescope set up on a very stable foundation with no vibration.   Balconies and rooftops are typically terrible—for both vibration and turbulence!

I run a 5” F/6.5 system in Chile under mostly very good seeing and I often get FWHM stars in the range of 1.4” - 1.6” when the seeing is good.  My feeling is that if you tune it up, you should be able to get your system FWHM down to at around 2” ( HFR < 2 px)  with good seeing.

John

Thank you John for the detailed answer. Always great to read your comments.

May I ask what is your method of choice for determining your FWHM in your subs? I see a massive discrepancy between what NINA Hocus focus reports, and what Pixinsight FWHMeccentricity script reports.
Like
macmade 3.01
Topic starter
...
· 
·  Share link
John Hayes:
The diffraction limited Airy diameter for your F/6.3 system is about 8 microns (7.68 um at 0.5 microns).  If you are measuring a 4.2 pixel half-flux radius with a 2.9 micron sensor, that’s a star diameter of roughly 24 microns, which is 5.3 arc-seconds in the sky.  That’s pretty big and I suspect that unless your seeing is really bad, that you should be able to do better.

There are a lot of factors that go into determining the star size in your integrated images.  First, you need high quality optics and they need to be very well aligned.  Second, you need perfect focus.  Your autofocus curve looks good so that shouldn’t be a problem. Third, you need excellent tracking and really good mechanical stability.  For your scope, if your total rms tracking errors are more than roughly 0.4”, under good conditions, that will be the limiting factor in determining your star size.  Finally, you need a very stable atmosphere.  The good news for you is that with a 6” scope, you should able to tolerate pretty mediocre conditions (around 2”) and still get pretty good results.

Do an experiment and take some very short exposures—less than say 10s.   Use bright stars if you have to and then measure the HFR.  That will tell you how well you autofocus and optical system are working independent of tracking errors.  Also check to make sure that you have the telescope set up on a very stable foundation with no vibration.   Balconies and rooftops are typically terrible—for both vibration and turbulence!

I run a 5” F/6.5 system in Chile under mostly very good seeing and I often get FWHM stars in the range of 1.4” - 1.6” when the seeing is good.  My feeling is that if you tune it up, you should be able to get your system FWHM down to at around 2” ( HFR < 2 px)  with good seeing.

John

Thank you so much, John, for your very precise answer!
I will experiment as you suggest and see what I can do from here. Thanks for the tips on how to proceed.

My setup is located on a balcony; there are indeed some vibrations.
It's really helpful to know what I should try to aim for!
Like
WhooptieDo 10.40
...
· 
·  1 like
·  Share link
HFR is a completely arbitrary number unfortunately.    I own two identical setups and get completely different HFR values.  I have a friend with a third identical setup and his HFR values are substantially different as well.  All scopes are getting the same FWHM readings.   The only thing HFR is good for is measuring relative conditions on your scope, and autofocusing. 

The only true way to measure sharpness is to drop a sub in Pix and measure FWHM.
Like
jhayes_tucson 26.84
...
· 
·  3 likes
·  Share link
Ashraf AbuSara:
May I ask what is your method of choice for determining your FWHM in your subs? I see a massive discrepancy between what NINA Hocus focus reports, and what Pixinsight FWHMeccentricity script reports.


Ashraf,
That's a good question!  My favorite way to measure FWHM is to use the tools in PixInsight.  The SubFrameSelector, FWHMEccentricity tools are excellent but they also require some care to make sure that you are getting the right results.  You have to make sure that you have the arcsec/px value set properly and it's especially important to make sure that the star detection parameters are set correctly.  For most cases, fit to Mofffat4 option and make sure that set the parameters to exclude overexposed stars and hot pixels.  When in doubt, you can confirm the results using the DynamicPSF PSF tool.  It will also show you the best fit Moffat order.   If you aren't careful with these tools it totally possible to generate random numbers!

I've found that the HFR measurements made by SGP aren't very good.  I don't know how NINA does but I hope that it's better than SGP!

John
Like
jhayes_tucson 26.84
...
· 
·  2 likes
·  Share link
Jean-David Gadina:
My setup is located on a balcony; there are indeed some vibrations.
It's really helpful to know what I should try to aim for!


Jean-David,
Yeah...imaging from a balcony is generally risky and not a good idea.  Here's something to keep in mind:  An arc-second is roughly a dime at 2 miles!  It is a VERY small angle and at that angle, you should think of everything as being made of rubber.  Any tiny building vibrations that will be present on your balcony will be quite a bit larger than 1 arc second!  My guess is that you'll immediately see a noticeable improvement in image quality if you move your scope to a more stable base on the ground.  Balconies are very convenient (I get it) but unfortunately that's not a good location for imaging.

Good luck with it!

John
Like
andreatax 9.89
...
· 
·  2 likes
·  Share link
There are balconies and balconies and when I was using one, very large and supported only on two sides, you could easily see visually (on the screen) the bending due to additional people standing on it. OTOH few of the other balconies I have used were rock solid, so here your mileage may vary. This said, the flexure didn't prevented me from taking pretty decent resolution pictures in the day. The main issue I suspect is that local environment may conspire against you as hot air induced turbolence is a pretty common phenomenon, in both buidling exteriors and balconies and can last for the most part of the night in summer. Worse of all are rootop terraces. If you have access to the NCP you can eliminate all other sources of errors due to tracking by shooting short exposures (from fraction of to few seconds) with the scope pointing at the NCP untracked and see how your average FWHM changes as you increase exposure length. Also internal turbolence due to temperature differential in SCTs is a common cause of reduced resolution, so watch out for those by observing an out-of-focus bright star and look if you see swirling patterns or teardrop/squashed shape of the Fresnel pattern of the out-of-fcous star. That an indication of internal currents in the OTA.
Like
macmade 3.01
Topic starter
...
· 
·  Share link
andrea tasselli:
There are balconies and balconies and when I was using one, very large and supported only on two sides, you could easily see visually (on the screen) the bending due to additional people standing on it. OTOH few of the other balconies I have used were rock solid, so here your mileage may vary. This said, the flexure didn't prevented me from taking pretty decent resolution pictures in the day. The main issue I suspect is that local environment may conspire against you as hot air induced turbolence is a pretty common phenomenon, in both buidling exteriors and balconies and can last for the most part of the night in summer. Worse of all are rootop terraces. If you have access to the NCP you can eliminate all other sources of errors due to tracking by shooting short exposures (from fraction of to few seconds) with the scope pointing at the NCP untracked and see how your average FWHM changes as you increase exposure length. Also internal turbolence due to temperature differential in SCTs is a common cause of reduced resolution, so watch out for those by observing an out-of-focus bright star and look if you see swirling patterns or teardrop/squashed shape of the Fresnel pattern of the out-of-fcous star. That an indication of internal currents in the OTA.

Thanks a lot!
Unfortunately the NCP is out of reach for me. I'll try experimenting with short exposures anyway, as @John Hayes recommended.
I'll also check internal turbulences as you suggest. That's very interesting!

My building is quite old, and the balcony is not very sturdy.
I can try using the Celestron vibration suppression pads, but judging by how they look, I'm not sure they will help.

I also think I need to try balancing my OTA better.

I'm looking forward to experimenting with all these recommendations and seeing how I can improve my shots!
Thanks again!
Like
McComiskey 3.01
...
· 
·  Share link
In terms of FWHM measurement, I ran some analysis of NINA/HF figures vs Pixinsight figures a couple of months ago.  I dove into that largely because NINA was telling my FWHMs were high versus the seeing monitor on site.

NINA's calculation of FWHM was consistently between 0.2" and 0.3" above the FWHM measured by Pixinsight.  I reached out to George Hilios, the developer of the HF module in NINA, and he knows this is the case, and explained that he uses a different methodology in selecting and measuring stars for this calculation than does Pixinsight.

I am not qualified to weigh in on the correctness of one approach versus the other, but ultimately concluded that it didn't matter much - I could use the NINA figures, at least on a relative basis, to tune the system performance.
Like
Gondola 8.11
...
· 
·  1 like
·  Share link
Vibration pads are worth a go. I once had a backyard location the was next to a large street. Every time a large truck came be I could see the ground vibrations through the scope. I made some DIY vibration pads and the problem went away.

On NINA HFR, I just use it to make sure I have best focus, which I do manually. I just tweak it until I get the lowest value. I have the same camera that you have and at an FL of 900mm, a good HFR would be around 2.6 to 2.9 The thing is, NINA measures HFR from the completed exposure. The longer your exposure time the larger the HFR value will become due to star bloat. That's one of the reasons I like shorter sub-exposures. Your camera has very low read noise as long as you set the gain to 252 or higher so you can have hundreds of subs without getting swamped with noise.
Like
 
Register or login to create to post a reply.