20-03-2021 Seeing-Resolution comparison in WL

this is the main message area for anything solar :)
Post Reply
H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

Dear all,

I hope you are well and have beautiful sunny days ahead of you.

I am still experimenting while making my first steps in solar photography, while my only available tools for the moment are my C8, the Astrosolar ND 3.8, ASI290, Baader Continuum (and a few more filters not used for the following photos) and the SSM by AiryLab to measure the seeing.

On the 20th of March I had the best seeing since I got the SSM to measure it: an average just below 1 arc second with minimum reaching 0.5 instantaneously! This is still not enough for the 208 mm aperture of the C8. I made many captures and below are 7 of them in a composite photo, presented together with the actual seeing. All captures were of 30 seconds that resulted in 4300-5000 frames each, of which I kept ONLY the BEST 20 for the stacking, in order to have only the frames during very minimum seeing number.
Seeing-Resolution comparison.jpg
Seeing-Resolution comparison.jpg (1.93 MiB) Viewed 240 times
I hope you can see the photo in full resolution so that comparisons between the 7 captures are possible (the violet horizontal bars are the 30 second periods of captures). I tried to rank the 7 captures according to their best minimum seeing and according to the real resolution (quality) I got. It seems to me (but please make your own ranking if you wish to play in this exercise) that the rankings are the following:

According to actual resolution (photos): 3, 4, 2, 1, 5, 7, 6
According to minimum seeing (SSM plot): 5, 4, 1, 3, 2, 6, 7

I would expect a better correlation than this. My ranking may not be objective, but I think you would all agree that capture 5 for example should be expected to have the highest resolution and it turned to be one of the worst.

Any comment is very welcome, especially if you have ideas to improve the experiment.

Best wishes,
Alexandros


User avatar
MAURITS
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 2131
Joined: Tue Nov 27, 2018 4:37 pm
Location: Belgium
Has thanked: 151 times
Been thanked: 343 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by MAURITS »

Interesting comparation Alexandros.


Regards,
Maurits

Vista del Cielo Observatory

www.vistadelcielo.be
User avatar
Carbon60
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 8651
Joined: Wed Mar 07, 2012 12:33 pm
Location: Lancashire, UK
Has thanked: 1682 times
Been thanked: 1523 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by Carbon60 »

Hi Alexandros,

The other factor is the exposure time per frame. Bright images with exposures of a few microseconds will give better results than less bright images with exposures of a few (tens of) milliseconds. Slow images will still suffer from blurring even under optimal seeing conditions. Please can you say what you used for these captures?

Stu.


H-alpha, WL and Ca II K imaging kit for various image scales.
Fluxgate Magnetometers (1s and 150s Cadence).
Radio meteor detector.
More images at http://www.flickr.com/photos/solarcarbon60/
User avatar
Montana
Librarian
Librarian
Posts: 24964
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 3769 times
Been thanked: 1497 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by Montana »

You have just discovered that great seeing isn't always correlated with image quality :)

Did you use a UV/IR cut filter? if not, I think this would clear up the fuzz :)

Alexandra


christian viladrich
Almost There...
Almost There...
Posts: 857
Joined: Sun Jun 14, 2015 4:46 pm
Location: France
Been thanked: 305 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by christian viladrich »

Hello,
With these conditions, you can improve image quality by using a red filter.
As said previously, it is better to keep exposure time short (< 4 to 5 ms).
A seeing scintillation monitor in sensitive to turbulence < about 2000 m. So the higher stage of turbulence (such as jet stream) is not detected.
Finally, the image resulting from the AS stack is not the addition of the "best frames" of the film. It is made of stacks of small pieces taken from different frames. Each stack is taken (generally) from difference frames.


Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Co-author of "Astronomie Solaire"
http://www.astronomiesolaire.com/
christian viladrich
Almost There...
Almost There...
Posts: 857
Joined: Sun Jun 14, 2015 4:46 pm
Location: France
Been thanked: 305 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by christian viladrich »

This being said, the SSM is very useful to check the evolution of seeing during the day, and also comparing different places.


Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Co-author of "Astronomie Solaire"
http://www.astronomiesolaire.com/
User avatar
Martin_S
Almost There...
Almost There...
Posts: 505
Joined: Sat Mar 16, 2019 12:34 am
Location: Brisbane , Australia
Has thanked: 144 times
Been thanked: 483 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by Martin_S »

Hi Alexandros, this is a great post, thank you for your excellent contribution. As Christian and Stu have stated it's best to keep exposure times short.
I don't have a SSM so I have taken a different approach with my WL Imaging.

I use a MAK180 with ND3.8 solar film and a selection of aperture masks to suit my seeing conditions.
I use a #58 green filter with an IR cut filter, this gives a faster exposure time than a Baader Solar Continuum filter.
For my captures I use the firecapture autorun feature.

200captures of 20 frames @ 1 second intervals. I then stack the best 2 of each batch and review them in Ifran view. I then save the best files for further
processing. This may be an unorthodox approach to WL imaging , I am constantly experimenting with different techniques to achieve the best images.

A fast PC with 16gb of RAM and a SSD really make a difference

Martin
WL_114314.txt
(750 Bytes) Downloaded 8 times


H - alpha setup
Lunt LS60THa Tilt-Tuned Solar Telescope, LS50FHa solar filter front-mounted Double stack etalon. 2 x Barlow
Remote etalon and focusing control's ASI290mm Camera

White Light setup
Skywatcher ED72 ,100 f/10 Achromat Homemade solar wedge, IR cut filter, #58 green filter.


AZEQ5 Pro mount.

Darkside setup

Skywatcher ED72AZGTi mount , GSO RC6. Celestron Evolution mount
User avatar
marktownley
Librarian
Librarian
Posts: 30737
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 4060 times
Been thanked: 1781 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by marktownley »

Thanks for this Alexandros. I'm not sure if you still have the original avi files, but I think you will get better results stacking more than 20, try maybe 100? This should get rid of more noise and should make ;signal' pop out more.

You can measure the quality of each of the 7 frames objectively; open each one as an individual image in AS3 and note the 'quality value' recorded, the higher this number the better.


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
christian viladrich
Almost There...
Almost There...
Posts: 857
Joined: Sun Jun 14, 2015 4:46 pm
Location: France
Been thanked: 305 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by christian viladrich »

Martin_S wrote: Mon Mar 22, 2021 6:50 pm
I use a #58 green filter with an IR cut filter, this gives a faster exposure time than a Baader Solar Continuum filter.

MartinWL_114314.txt
Indeed, the continuum filter is a bit "dark" when used with density 3.8. With the W58 (or with a trichromatic green filter), you get more light and it is easier to shorten the exposure time. It is worth experimenting for sure.


Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Co-author of "Astronomie Solaire"
http://www.astronomiesolaire.com/
H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

Dear all,

Thank you so much for your comments and input! Lot's of ideas for more experimenting and improvement! :-)

@ Stuart: I always try to be very close or just below 1 ms exposure with the Continuum. In this case the shutter was at 1.040 ms. In a previous post by Martin, Martin thanked you Stuart for some precious advice from your part that greatly improved his WL images. Do you remember what was your 'magic' advice? :-)

@ Alexandra: Following precious advice by Marty, I bought the IR pass 685 nm that allows me exposures of 0.320 ms(!) and the H-a 656 nm (12 nm) that allows only for slow exposures of some 3.4 ms.

Despite the differences in these filters and also the long exposure of H-a, so far, my comparisons did not show clearly the filter that gives the best results for WL. It seems that the variation of seeing is more important than the filter? Is it because I never had so good seeing that would make the comparisons meaningful? The same day with the above photos with similar conditions of seeing I used IR pass 685 and and H-a and the intra-filter variation is similar to the intra-capture variation... I will repeat the tests, but if someone has already made them in the past, it would be great to see what results they gave. Ideally, we should use simultaneously two or three telescopes to eliminate the seeing variation factor, but I don't have this option...

@ Christian: Cher Christian c'est un grand honneur!! So far I had the honor to receive comments and advice from many experienced and kind people here. Since you have used Celestron scopes for solar imaging a lot, before entering the chat and having read many of your posts and articles and having observed your images and tests here and elsewhere I was planning to email you as I was desperate to find a way to use my C8 for solar imaging... Then I said, better to try to understand the very basic things and then try to contact you. The 'red filter' you mentioned would work better than the IR pass 685 or the H-a 656? Given that I keep my exposure as fast as possible, apart going to the mountains (I will do so :-) as soon as our Greek lock-down ends or I get a paper to legally do so...), what else could improve the images of C8?

Regarding the process method, I let PIPP select the 500 best out of the 5000, and then stack the 20 best according to AS!3.
Thanks for the explanations about the SSM seeing measurements! Precious info.

@ Martin: Is masking improving the resolution of your results? I currently have a 180mm Astrosolar ND 3.8. If I mask it further I could have better results when the seeing is not perfect? Thanks for the filter idea. I will try it!
"200captures of 20 frames @ 1 second intervals. I then stack the best 2 of each batch". Very interesting. What is the idea behind this way of capturing/stacking? Why it should be better in theory?

@ Mark: "You can measure the quality of each of the 7 frames objectively". Excellent! Thanks so much! This will improve any future experiment. I really needed this! :-)

I just tried and if the number you meant is the 'Q', I got:

Photo Quality number AS!3 Rank 'My eye Rank'

1 1317404 1 4
2 743821 5 3
3 871556 4 1
4 1302047 3 2
5 657937 6 5
6 1316983 2 7
7 350430 7 6

IT IS ANOTHER WORLD! :-( I have to have my eyes checked! ;-) How is the Q estimated? On which criteria?
For me 3 and 4 were clearly the best and 6 and 7 the worst. AS!3 does not agree at all! :-(


Regarding number of frames to stack, I am puzzled... I tried 100 and some times it is better, but some times it is much worst! I really do not understand how it works... :-( I also experimented with selecting more than 500 (through PIPP) from the original 5000. Once more, some times the results are better, but some are worst... :-( I suppose you have all experimented a lot and ended up with your way of analyzing the captures/frames.

Thanks once more to all of you!
Alexandros
Last edited by H-Alpha on Tue Mar 23, 2021 12:42 am, edited 1 time in total.


User avatar
marktownley
Librarian
Librarian
Posts: 30737
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 4060 times
Been thanked: 1781 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by marktownley »

Meant to say, check the Q value on the stacked, not post processed images, then post process the one with the highest value.

Not sure exactly how it works, but seems to be a measure of how wide, or how much, of the histogram is used. In images where seeing is poorer the histogram isn't as wide, whereas for the same feature more of the histogram is used.


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

marktownley wrote: Mon Mar 22, 2021 10:19 pm Meant to say, check the Q value on the stacked, not post processed images, then post process the one with the highest value.

Not sure exactly how it works, but seems to be a measure of how wide, or how much, of the histogram is used. In images where seeing is poorer the histogram isn't as wide, whereas for the same feature more of the histogram is used.
You are right Mark. I did it on the processed images... I will repeat it to see the differences.
If it measures the histogram, I hope it is not affected by the width of the original histogram during capture, because I use very low gains and no Gamma, so my histograms are short... Is this wrong?

Hope it measures within the available histogram somehow, by weighting between histograms of various original width.


H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

I try it and interestingly it is almost identical to the ranking of the processed images!!!

Photo --- AS!3 Ranking orginal - AS!3 Ranking processed --- My visual ranking

1 --- 2 - 1 --- 4
2 --- 5 - 5 --- 3
3 --- 4 - 4 --- 1
4 --- 3 - 3 --- 5
5 --- 6 - 6 --- 7
6 --- 1 - 2 --- 7
7 --- 7 - 7 --- 6


H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

Montana wrote: Mon Mar 22, 2021 2:29 pm You have just discovered that great seeing isn't always correlated with image quality :)

Did you use a UV/IR cut filter? if not, I think this would clear up the fuzz :)

Alexandra
Hi Alexandra,

I checked the spectra of the filters I use (below). The Continuum (540 nm) and the H-Alpha 656 nm are very selective and short. According to the spectra of Baader and Astronomik they should not let UV/IR pass.

Should I add an UV/IR cut as you propose anyway, because the spectra may not be correct?
Attachments
astronomik-ha12-ccd_trans.png
astronomik-ha12-ccd_trans.png (57.61 KiB) Viewed 132 times
baader-solar-continuum-filter-1-1-4-540nm-dea.jpg
baader-solar-continuum-filter-1-1-4-540nm-dea.jpg (76.29 KiB) Viewed 132 times


H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

Carbon60 wrote: Mon Mar 22, 2021 7:20 am Hi Alexandros,

Bright images with exposures of a few microseconds will give better results than less bright images with exposures of a few (tens of) milliseconds. Slow images will still suffer from blurring even under optimal seeing conditions.
Stu.
Stuart can you please explain this a bit more? I try to use the shorter exposure possible (~1millisecond or less) depending on the filter. How to achieve few microseconds? Do you mean tenths of milliseconds or microseconds? (1 millisecond = 1000 microsecond, isn't it?)
Is a 1-millisecond exposure a slow one that creates blurring?


H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

Martin_S wrote: Mon Mar 22, 2021 6:50 pm
200captures of 20 frames @ 1 second intervals. I then stack the best 2 of each batch and review them in Ifran view. I then save the best files for further processing.
Martin, by 1 second interval you mean 1 second delay between each of the 200 autorun captures. Isn't it?

And then you stack just two (just the two best?) frames out of each 20 frames? So you end up with 200 images to view and select for further processing? You go visually through 200 images? Probably, I did not understand it right... :-( Can you please help me to understand?


User avatar
Martin_S
Almost There...
Almost There...
Posts: 505
Joined: Sat Mar 16, 2019 12:34 am
Location: Brisbane , Australia
Has thanked: 144 times
Been thanked: 483 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by Martin_S »

H-Alpha wrote: Tue Mar 23, 2021 1:15 am artin, by 1 second interval you mean 1 second delay between each of the 200 autorun captures. Isn't it?

And then you stack just two (just the two best?) frames out of each 20 frames? So you end up with 200 images to view and select for further processing? You go visually through 200 images? Probably, I did not understand it right... :-( Can you please help me to understand?

That's exactly what I do for now and I'm getting good results, I found that if I stack a lot of bad frames from say one video file the stacked image is not so good as one or two good stacked frames. I try to minimise the number of bad frames.


H - alpha setup
Lunt LS60THa Tilt-Tuned Solar Telescope, LS50FHa solar filter front-mounted Double stack etalon. 2 x Barlow
Remote etalon and focusing control's ASI290mm Camera

White Light setup
Skywatcher ED72 ,100 f/10 Achromat Homemade solar wedge, IR cut filter, #58 green filter.


AZEQ5 Pro mount.

Darkside setup

Skywatcher ED72AZGTi mount , GSO RC6. Celestron Evolution mount
User avatar
Carbon60
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 8651
Joined: Wed Mar 07, 2012 12:33 pm
Location: Lancashire, UK
Has thanked: 1682 times
Been thanked: 1523 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by Carbon60 »

H-Alpha wrote: Tue Mar 23, 2021 12:36 am
Carbon60 wrote: Mon Mar 22, 2021 7:20 am Hi Alexandros,

Bright images with exposures of a few microseconds will give better results than less bright images with exposures of a few (tens of) milliseconds. Slow images will still suffer from blurring even under optimal seeing conditions.
Stu.
Stuart can you please explain this a bit more? I try to use the shorter exposure possible (~1millisecond or less) depending on the filter. How to achieve few microseconds? Do you mean tenths of milliseconds or microseconds? (1 millisecond = 1000 microsecond, isn't it?)
Is a 1-millisecond exposure a slow one that creates blurring?
Hi Alexandros,

Sure.... not knowing your actual exposure times (per frame), I was making a general statement about this additional factor that has to be considered when making high definition images. In my case I use a Lacerta wedge with the neutral density filter stripped out and either a green or red filter set in its place. I also have a UV/IR filter on the nose of the wedge. This produces very bright images which I then attenuate using one or two polarising filters on the nose of the camera or Barlow (strictly for imaging, not visual)

Rotating the polarisers gives me full control of the image brightness and, consequently, the exposure time. In practice I can take things to the max and get down to 100-200 microseconds, or less (based on what my imaging software tells me....I don’t know if this is the actual speed), or, by turning the polarisers, up to several tens of milliseconds. Yes 1 millisecond is 1000 microseconds.

At the image scale you’re using in this demo I would have thought 1 millisecond would have been sufficient to freeze the action, but as has been said, the SSM isn’t reliable in all cases. It depends on the nature of the turbulent air.

Stu.


H-alpha, WL and Ca II K imaging kit for various image scales.
Fluxgate Magnetometers (1s and 150s Cadence).
Radio meteor detector.
More images at http://www.flickr.com/photos/solarcarbon60/
User avatar
marktownley
Librarian
Librarian
Posts: 30737
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 4060 times
Been thanked: 1781 times
Contact:

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by marktownley »

Are you collimating the C8 regularly?

1" seeing at 540nm corresponds to ~135mm aperture critically sampled, 0.5" corresponds to ~270mm aperture. I know those are the values the SSM are telling you, but take a look at the full size version of this image posted on the forum a few days ago viewtopic.php?f=4&t=31145 This is with 110mm and you can see whilst your granulation cells are larger as you would expect with the longer focal length, those with the smaller aperture are resolved better.

This says to me your seeing isn't quite as good as the SSM might suggest, or collimation, or both...


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
H-Alpha
Oh, I get it now!
Oh, I get it now!
Posts: 43
Joined: Sun Feb 21, 2021 2:19 am
Has thanked: 79 times
Been thanked: 33 times

Re: 20-03-2021 Seeing-Resolution comparison in WL

Post by H-Alpha »

marktownley wrote: Tue Mar 23, 2021 8:03 pm Are you collimating the C8 regularly?

1" seeing at 540nm corresponds to ~135mm aperture critically sampled, 0.5" corresponds to ~270mm aperture. I know those are the values the SSM are telling you, but take a look at the full size version of this image posted on the forum a few days ago viewtopic.php?f=4&t=31145 This is with 110mm and you can see whilst your granulation cells are larger as you would expect with the longer focal length, those with the smaller aperture are resolved better.

This says to me your seeing isn't quite as good as the SSM might suggest, or collimation, or both...
Thanks Mark.

After reading all the comments and looking at photos of other people, I started to wonder what i am doing wrong... So you are right to wonder about collimation. Since 1990 that I bought the C8 and until one month ago, I did not even know the term "collimation". When i realized how important it is, I collimated for the first time and surprisingly, I saw that the C8 was very slightly off collimation, so I corrected it during the night with a bright star. To my eyes it seemed perfect. I will try once more. Are doing this with laser and special equipment or just with a star? And how often should I check the scope? Every week or month? Is it possible to check if collimated during the day?

Mark can you please tell me which are the rules/formulas to estimate the the aperture for a given Seeing and a given frequency in the spectrum?

I was a bit confused about the filters proposed to try at the end... I have the L of my LRGB Zwo filters to use as UV/IR cut. Can it work in combination with the Baader IR pass (685 nm) that gives me bright images, or should I use it just with the ZWO Green or Red?

And something last: Masking my C8 could improve my resolution when the seeing is not appropriate or not?


Post Reply