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

Trying to diagnose my imaging resolution limiters: Did an experiment

  • Please log in to reply
7 replies to this topic

#1 smiller

smiller

    Vostok 1

  • *****
  • topic starter
  • Posts: 133
  • Joined: 27 Oct 2018

Posted 25 November 2021 - 08:16 PM

Hi Folks,

 

This is a bit of a long analysis, my apologies if it exceeds your patience...

 

I've been out imaging Jupiter three times over the last 3 weeks, each time doing 10-20 captures, and I've had remarkable stable results each night.   

 

I've played with image scale (From ~0.13" to 0.25"), improved my focusing and scope stability, played with biasing focusing just a tiny touch one way or another, optimized frame rates on FireCapture, started to fine tune my ADC use, and played with a zillion options optimizing enhancements in stacking and sharpening.

 

But the results come back very very similar, as if they are capped for the maximum quality image I get in an evening.  I can sometimes get a bad capture (poor focus, haze or cloud comes over, or seeing just tanks for a couple of minutes, etc....) but 80% of my captures, after processing, come out almost exactly the same.  So that makes me wonder:

 

1) Either I'm seeing limited and seeing happened to be pretty much the same all three evenings and generally very consistent within an evening (minus a cloud or two).

 

2) Or my 12" Dob has a pretty hard limit (ex: Collimation or other optical issue).   For instance, I've been using a collimation cap and only some minimal star tests so I know it's just basically collimated and could be better.

 

3) Or my technique for simulating poor seeing is... um... poor substandard.

 

So I did an experiment.  If my scope isn't the issue, than if I artificially degrade seeing from my best captures, I should immediately see a clear difference.  If my scope IS the limit then I shouldn't lose much image quality because there are tons of frames that are just scope limited.

 

So I did it in three ways for a good 3 minute ~16,000 frame capture of Jupiter at 0.13" resolution:

 

1) I used PIPP to take only every other, third, fourth, or even fifth frame.  I then stacked 2700.  The theory is frames are tossed out consistently and at 1/3, I only have 1/3 the number of the best frames, so I'm clearly tossing lots of more mediocre ones in there.

 

2) I looked at Autostakkert's analysis and then used PIPP to take a segment the of video that had the worst seeing.  Almost all the frames were rated by Autostakkert to be from 5% to 60%, averaging about 40%

 

3) I had PIPP Sort the frames and then stored only about 3200 of the worst frames.

 

I stacked 2700 of these using the same options and alignment points. 

I used Registax and used the same routine/parameters, but I noticed on a couple i needed to adjust just a tad to try to get the equivalent balance of sharpness/noise.  In general, I tried to get the best out of each, striving for a reasonable amount of punch to see what popped out.  I didn't take forever to make them equivalent, just looking for the big picture.

 

In looking at the videos, I could see differences, they weren't dramatic, but I expected a clear difference in result.

 

Here are the results:

 

The original File of 16000 frames with the best 2700 stacked:

 

F1AllconvSharp4Less.jpg

 

Taking every third capture, so only 1/3 the amount of frames to choose the best from:

 

F1AllconvSharp4Less.jpg

 

With only 1/5 the number of frames, the quality then finally started to roll off (not pictured but similar to the next two).

 

 

Now using the worst 3200 frame continues sequence of frames as reported by Autostakkert's analysis:

 

TheWorstSeeingsequenceconvRegi.jpg

 

And finally, the worst frames as reported by PIPP (with 80% under 50% and average about 35%)

 

SortedForWorstconvRegi.jpg

 

 

So not a huge difference.  What does your judgement and experience tell you, that I have some pretty hard limit in my imaging system?

 

Other than just getting a selection of alignment tools and whacking at it, is there a way to diagnose what the cause may be, if there is a cause other than seeing?  I feel like an auto mechanic and I'm going to be tuning things, changing out things, and I'm not even sure if the car is the issue.

 

For the record, I have a FarPoint 2" Cheshire on order so I'm starting to up my collimation tool set and will work on my procedure.   As a blind mechanic, I'm guessing this is one potential culprit... and in any case even if not now iIm told it will quickly become the limiter.

 

 

Happy Thanksgiving!

 

Steven

Attached Thumbnails

  • F3ThirdconvRegi.jpg

Edited by smiller, 25 November 2021 - 09:02 PM.


#2 JMP

JMP

    Apollo

  • *****
  • Posts: 1,224
  • Joined: 31 Oct 2005

Posted 26 November 2021 - 10:43 PM

Hi Steven; You can test collimation by imaging stars, some doubles are interesting. If your scope is collimated the diffraction rings will be symmetrical.

When my images seem off it frequently is the collimation.
  • smiller likes this

#3 smiller

smiller

    Vostok 1

  • *****
  • topic starter
  • Posts: 133
  • Joined: 27 Oct 2018

Posted 26 November 2021 - 10:59 PM

Hi Steven; You can test collimation by imaging stars, some doubles are interesting. If your scope is collimated the diffraction rings will be symmetrical.

When my images seem off it frequently is the collimation.

I initially missed that you said "imaging stars" not viewing.  Do you mean with the method used for planets?  (I've never done stars).  Wow, great idea.

 

Roughly how much exposure should I use (relative to Jupiter for instance) to not blow out the image with star bloom but to see the rings?

 

Cheers,

 

Steven


Edited by smiller, Yesterday, 10:19 AM.


#4 RedLionNJ

RedLionNJ

    Aurora

  • *****
  • Posts: 4,748
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted Yesterday, 09:47 AM

For twelve inches of aperture, these are all falling somewhere on the mediocre/so-so scale.  It's most likely you're at the mercy of the seeing, but while somebody may rate seeing as "3/5" on a particular night, one person's 3/5 is not going to be identical to another's.  There are lots of different types of mediocre or poor seeing and they will almost always lead to mediocre or poor results. It's been said a thousand times and bears repeating yet again - seeing is king.

 

These experiments would be best carried out under really good seeing. But if you don't get really good seeing very often, you're going to want to make the most of it for actual imaging. That's the quandary I keep running into.

 

JMP's suggestion of using double stars to rule out your scope is an excellent one. You can pick targets high in the sky (certainly not lurking in the murk like Jupiter & Saturn right now) and take your time. The exposure, etc, will depend on the brightness of the stars, but I usually try to find 5th or 6th magnitude pairs, as finding a nice gain/exposure combination seems very easy on those.

 

It's possible your collimation is a bit off, but I'm also 99% sure you're seeing-limited in your experiments above.


  • smiller likes this

#5 smiller

smiller

    Vostok 1

  • *****
  • topic starter
  • Posts: 133
  • Joined: 27 Oct 2018

Posted Yesterday, 10:24 AM

For twelve inches of aperture, these are all falling somewhere on the mediocre/so-so scale.  It's most likely you're at the mercy of the seeing, but while somebody may rate seeing as "3/5" on a particular night, one person's 3/5 is not going to be identical to another's.  There are lots of different types of mediocre or poor seeing and they will almost always lead to mediocre or poor results. It's been said a thousand times and bears repeating yet again - seeing is king.

 

It's possible your collimation is a bit off, but I'm also 99% sure you're seeing-limited in your experiments above.

Thanks for sharing your wisdom.  I guess I was being hopeful that I was in control of this and not the seeing gods.  I really just need a lot more opportunities to find that one magic evening, but I’m in the rainy Pacific NW so have limited opportunities in the winter months.  Patience… patience…. Oh well…. Thank goodness for “cloudynights”.



#6 JMP

JMP

    Apollo

  • *****
  • Posts: 1,224
  • Joined: 31 Oct 2005

Posted Yesterday, 12:58 PM

Hi Steven, I'm also in the NW, near Eugene, OR. We do get a fair number of nights with 4/5 seeing according to Clear Sky Chart. Many times though the best seeing is after midnight.

Some of the best seeing can be found on tropical islands, Damian Peach has two C14's, he keeps one in Barbados in the Caribbean.

Here's an illustration of how critical collimation can be, the double-double is almost overhead on summer nights. With my C11 one double is well collimated, on the other side of the chip the other double is slightly off:

280-Epsilon-Lyrae_C11.jpg
  • LauraMS and smiller like this

#7 ch-viladrich

ch-viladrich

    Viking 1

  • -----
  • Posts: 787
  • Joined: 14 Jul 2013
  • Loc: France

Posted Today, 01:58 PM

I understand you have a 12" Dob ?

 

My recommendations would be :

- improve your collimation skills. Basic collimation is not enough for HR collimation.

- spend a lot of time star testing : you improve for hability for collimation, you will identify the limits of your optics (are you sure your optic is good? is it well supported in its cell ?), we will learn to evaluate the seeing condition (including instrumental turbulence).

This is nothing difficult about it, but there is a learning curve. Books like "Star testing telecope / A. Suiter" can speed up the whole learing curve.

In any case, experimenting and testing like you did is the right way to do.


  • smiller likes this

#8 Ittaku

Ittaku

    Surveyor 1

  • -----
  • Posts: 1,818
  • Joined: 09 Aug 2020
  • Loc: Melbourne, Australia

Posted Today, 04:09 PM

Collimation. Ruthless attention to collimation makes an enormous difference at planetary scales, and by your own admission you only do basic collimation.


  • smiller likes this


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