Does this appear to be overprocessed?

this is the main message area for anything solar :)
Post Reply
Bruce G
Ohhhhhh My!
Ohhhhhh My!
Posts: 177
Joined: Mon Jul 16, 2018 8:18 pm
Has thanked: 128 times
Been thanked: 121 times

Does this appear to be overprocessed?

Post by Bruce G »

I was working on some images of our little spot. The seeing was pretty poor, as is often the case for my location, but I was trying to squeeze out detail and I got to a point where nothing looks right to me now. Is the finer detail here merely processing artifacts? Please enlarge the image to full size or even 200%.

The second image is what I started from.

Thanks for your comments

Bruce G

2020_08_07-11-45-57_Frame_693.jpg
2020_08_07-11-45-57_Frame_693.jpg (3.73 MiB) Viewed 896 times

2020_08_08-11-45-57_Frame_693 original.jpg
2020_08_08-11-45-57_Frame_693 original.jpg (950.05 KiB) Viewed 896 times


User avatar
dhkaiser
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 1702
Joined: Sat Dec 14, 2019 2:57 pm
Location: Indiana, USA
Has thanked: 1407 times
Been thanked: 899 times

Re: Does this appear to be overprocessed?

Post by dhkaiser »

Looks good to me!


Dan
South central Indiana
USA
User avatar
AJamesB
Im an EXPERT!
Im an EXPERT!
Posts: 372
Joined: Thu Jan 30, 2020 9:21 pm
Location: Maui, Hawaii
Has thanked: 1150 times
Been thanked: 630 times

Re: Does this appear to be overprocessed?

Post by AJamesB »

Looks pretty good! If I pixel peep, I do see some patterns/striations/structures I've not seen before in solar images, even very high resolution ones, almost looks like it could be false detail generated by a sharpening algorithm. I'll get something not too dissimilar if I try and use something like Topaz sharpen AI, but I'm so new to this that I would wait for others with more experience to weigh in on that aspect.

Contrast looks pretty good too, though (at least on my monitor) it looks like the blacks may have been crushed just a bit. For example, if you look at the very center of the core of the sunpot (the umbra I think?) in the unprocessed image, you can see a slightly lighter spot almost dead center of it. This gets lost though in the final image where the whole sun spot core is darkened to solid black (again, at least on my monitor, ignore if you can still see it!), which can come from too much contrast enhancement or something similar.

All in all though it looks good! Noise is almost non-existent and it looks very clean, especially when viewing unzoomed. I like it!


Lunt ls100tha single stacked + hinode solar guider on ioptron cem26 mount
asi178mm
asi294mm
2x, 3x, and 4x telecentrics
0.6x and 0.4x telecompressors
Astrophil
Almost There...
Almost There...
Posts: 903
Joined: Thu Dec 20, 2018 2:00 am
Been thanked: 581 times

Re: Does this appear to be overprocessed?

Post by Astrophil »

Hi Bruce,
I see what you mean especially around the smaller filament. I'm curious as to the scope used and the processing filter applied to this image. It sure seems like unusual details, at least it's nothing I could capture.

Phil


User avatar
rsfoto
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 6161
Joined: Mon Jun 18, 2012 8:30 pm
Location: San Luis Potosi, México
Has thanked: 9411 times
Been thanked: 5571 times

Re: Does this appear to be overprocessed?

Post by rsfoto »

Hi,

It is difficult to say when not having the original video. Everybody processes in a different way.

The only thing I see in the first image that is looks Sandy and has some strange filigree aspect.

The non stretched image (second image) has 1730x1087 pixels and the first image has 2592x1628 pixels. Why did you resize it ?

I downloaded your second image and give it a little stretch. The result tells me that the raw material = acquired video is not up to what you expected or the original video processing did not produce what you expected.

I always tell my Photography students. " From Marmalade you can make Poo but from Poo you can not make Marmalade " :lol: :lol: :lol: :lol:

If you care to share the video I would gladly look into it.

This IMHO ...

2020_08_08-11-45-57_Frame_693-original_et.jpg
2020_08_08-11-45-57_Frame_693-original_et.jpg (1.6 MiB) Viewed 879 times
BTW, Upsizing an image never improves quality. You always will have a quality loss. Downsizing makes every image look better

2020_08_08-11-45-57_Frame_693-original_down.jpg
2020_08_08-11-45-57_Frame_693-original_down.jpg (434.74 KiB) Viewed 879 times


regards Rainer

Observatorio Real de 14
San Luis Potosi Mexico

North 22° West 101°
Bruce G
Ohhhhhh My!
Ohhhhhh My!
Posts: 177
Joined: Mon Jul 16, 2018 8:18 pm
Has thanked: 128 times
Been thanked: 121 times

Re: Does this appear to be overprocessed?

Post by Bruce G »

Thanks for your comments folks.

ammonthenephite and Astrophil - Yes, you're seeing what I'm talking about. Interesting that you should mention Topaz products, ammonthenephite. I used Topaz Gigapixel AI for the deconvolution. Aside from that though, it's all contrast and brightness adjustments in Photoshop. No imPPG this time.
rsfoto - The image was upsized in order to apply noise filtering (I use Neat Image). Upsizing lowers the noise frequency. Also, Topaz Gigapixel AI enlarges by default. I should have resized it back down again when I made the final image. You are correct, the original is not stellar. The seeing was poor.

There is no need to post a video. The image I posted is produced entirely from the single frame that I posted (second image). Note the time stamp at the top. I wanted to avoid stacking artifacts and the stacks were really no better than the single frame (though they were lower in noise).


User avatar
Rusted
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 1738
Joined: Fri Jan 19, 2018 11:00 am
Location: Central Denmark
Has thanked: 8016 times
Been thanked: 1938 times
Contact:

Re: Does this appear to be overprocessed?

Post by Rusted »

Dropping your unprocessed image into iMPPG immediately produces a coarse grain, like sand.

Image processing is a minefield and what others see is entirely dependent on the computer monitor and the image hosting service [forum/
blog/etc.]

I often think I should produce multiple variations of my own astro "daubs" for posting on different forums and blogs. The differences in contrast and detail, between the resulting images, as seen on exactly the same monitors, is simply quite unbelievable!

In a worst case scenario one might think a solar/astro imager might have captured a unique and scientifically "interesting" feature. Yet, on a random scan of all the online hosts the feature would be completely invisible. Or so softly rendered as to be completely unrecognisable.

Perhaps the answer is to host one's images online where quality really matters and then link to them. This is something I usually dismiss as completely losing the immediacy of the forum image hosting system. Coming from over half a century of film cameras, I still find it magical that I can post an image to the world at large within seconds of capture.

My dissatisfaction with image quality has nothing to do with seeing conditions, image resizing or anything else. It is all about resolution and contrast of the hosted images. Now I keep wondering how one could cheat the system somehow. Other than radically over-processing every single image for forum hosting. The bottom line is that I simply don't recognise the images I post and over which I have slaved to squeeze out every detail without over-sharpening.


http://fullerscopes.blogspot.dk/

H-alpha: Baader 160mm D-ERF, iStar 150/10 H-alpha objective, 2" Baader 35nm H-a, 2" Beloptik KG3,
Lunt 60MT etalon, Lunt B1200S2 BF, Assorted T-S GPCs or 2x "Shorty" Barlow, ZWO ASI174.
User avatar
marktownley
Librarian
Librarian
Posts: 42272
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20435 times
Been thanked: 10245 times
Contact:

Re: Does this appear to be overprocessed?

Post by marktownley »

Hi Bruce,

When you look at the processed image full size it is full of artifacts.

If you look at the penumbral fibrils around the sunspot they're not resolved at all. This says to me seeing is causing blur. However look at various parts of the image and you see a strange pattern, like rockwool, or how fungal spores grow fine hairs - these are completely artifacts generated by the sharpening routine, definitely not real features.

You should have a go at stacking and seeing what the processed image looks like. What scope did you use?

Mark


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
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 2150
Joined: Sun Jun 14, 2015 4:46 pm
Location: France
Has thanked: 1 time
Been thanked: 2713 times
Contact:

Re: Does this appear to be overprocessed?

Post by christian viladrich »

Regarding the difference between image and posted image, it can be reduced by saving the image with a sRGB color profile.
Unfortunately, not all browsers use color profiles which can lead to differences in contrast, color or luminosity.
On top of that, monitors have their own biais. It could be a good idea to use a color probe to tune the color correction of the monitor.


Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
User avatar
yltansg
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 2798
Joined: Sat Jun 14, 2014 9:33 am
Location: Singapore (1.3N, 103.7E)
Has thanked: 1446 times
Been thanked: 1423 times

Re: Does this appear to be overprocessed?

Post by yltansg »

Hi Bruce,

Your single frame capture looks ok. I would like to have a go with your video.

Alfred


User avatar
Montana
Librarian
Librarian
Posts: 34563
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17672 times
Been thanked: 8790 times

Re: Does this appear to be overprocessed?

Post by Montana »

Enough said already :) it is a lovely pic though isn't it :)

Alexandra


User avatar
rsfoto
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 6161
Joined: Mon Jun 18, 2012 8:30 pm
Location: San Luis Potosi, México
Has thanked: 9411 times
Been thanked: 5571 times

Re: Does this appear to be overprocessed?

Post by rsfoto »

Bruce G wrote: Tue Aug 11, 2020 3:08 am Thanks for your comments folks.

ammonthenephite and Astrophil - Yes, you're seeing what I'm talking about. Interesting that you should mention Topaz products, ammonthenephite. I used Topaz Gigapixel AI for the deconvolution. Aside from that though, it's all contrast and brightness adjustments in Photoshop. No imPPG this time.
rsfoto - The image was upsized in order to apply noise filtering (I use Neat Image). Upsizing lowers the noise frequency. Also, Topaz Gigapixel AI enlarges by default. I should have resized it back down again when I made the final image. You are correct, the original is not stellar. The seeing was poor.

There is no need to post a video. The image I posted is produced entirely from the single frame that I posted (second image). Note the time stamp at the top. I wanted to avoid stacking artifacts and the stacks were really no better than the single frame (though they were lower in noise).
Hi Bruce,
There is no need to post a video.
It is not about posting a video here.

First thing is to get the best adjustment of the real time image on your monitor. The most important factor, at least for me, is getting the best focus which is an Art itself and takes quite some time. Sometimes it take sme much more time to get focus then acquiring a single video or the full session of 9 videos for my full discs.

One single image of a video can never be better then a stack of many images.

IMHO sometimes using too much frames for stacking will give a worse result then using less frames in stacking procedure.

Some of my Full disc have only 120 frames for stacking. OK perhaps I live in a place with good atmospheric conditions. Somebody from Australia when analyzing the SDE ( Sub Divisional Error) of my mounts in regard to the fitted encoders told me to move to a place with a worse seeing and so SDE would not influence my DSO imaging :lol: :lol: :lol: :lol: :lol:

Again, I insist, The better the raw material, the better the result will be ... and our way of imaging the Sun with the Lucky Imaging procedure is the best start e.g. take the best bits of each frame and assemble that Jigsaw Puzzle to a good one. The resulting Puzzle with all its pieces will be as good as each single piece is and not better.

OK, applying a bit of contrast, brightness and sharpening gives us a nicer to look at image.

IMHO


regards Rainer

Observatorio Real de 14
San Luis Potosi Mexico

North 22° West 101°
Bruce G
Ohhhhhh My!
Ohhhhhh My!
Posts: 177
Joined: Mon Jul 16, 2018 8:18 pm
Has thanked: 128 times
Been thanked: 121 times

Re: Does this appear to be overprocessed?

Post by Bruce G »

tusted - The sand grains, or dots, from the Lucy-Richardson method are a reason I was looking at a different deconvolution method.
Regarding the quality of posted images - what they do to my images on Facebook when I post them for friends is almost criminal. I want to cry sometimes and have even resorted to posting the image at different levels of magnification so that they can get an idea of the detail that I see. Telescopes are new to me. I have spent most of my years with microscopes and macro subjects, notably snowflakes. The beautiful detail in snowflakes astounds me at times (see below) and seeing it turned into mush is heartbreaking. I too am from the film era and both marvel at and whine about what digital has brought us.

Mark - I think you put your finger on what was disturbing me so much about the image. There was stuff that seemed too good to be true while other places lacked a similar apparent sharpness. We know that there should be penumbral detail and it's not there, so clearly anything finer than that has to be garbage. I guess the "fungus" is this program's version of the Lucy-Richardson dots

The scope is a Lunt LS80, single stacked
5x TeleVue Powermate
ZWO ASI174MM camera
Two arc seconds should be achievable

Alfred - It's a 4.6 GB .ser file and honestly, the quality isn't worth it unless you are simply up for a challenge and want to play with it. If you have a Dropbox or some other means of transferring a large file, I'll be happy to give it to you.

rsfoto - There is no question that good focus is the starting point. It becomes very frustrating if the camera does not have a good frame rate. I find that rather than looking at large objects, like the sunspot in this case, I can achieve more critical focus by looking at the small stuff. And I don't even look at a small detail, but rather I attempt to visually take in as much of the image as possible. Stacking gives improved signal to noise (proportional to the square root of the number of images stacked) and the also allows to push Mr. Nyquist a bit in terms of resolution because of the jittered sampling due to atmospheric disturbances. But as you state, the better the raw material, the better the result.

I usually shoot 1000 frame sequences and stack 5. 10 and 25%. I rarely use anything but the 5%, but sometimes 10% (100 images) provides some improvement

Thank you Alexandra!

Snowflake Center new1.jpg
Snowflake Center new1.jpg (1.35 MiB) Viewed 806 times
Dendrite1.jpg
Dendrite1.jpg (3.64 MiB) Viewed 806 times


User avatar
rsfoto
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 6161
Joined: Mon Jun 18, 2012 8:30 pm
Location: San Luis Potosi, México
Has thanked: 9411 times
Been thanked: 5571 times

Re: Does this appear to be overprocessed?

Post by rsfoto »

Alfred - It's a 4.6 GB .ser file and honestly, the quality isn't worth it unless you are simply up for a challenge and want to play with it. If you have a Dropbox or some other means of transferring a large file, I'll be happy to give it to you.
Hi Bruce,

I would like to look at your video. Send the Dropbox link for Downloading it. Thanks.

:bow2

It becomes very frustrating if the camera does not have a good frame rate.
Yes, that is what we all want but I have an oldie camera and an oldie video acquiring software and I only get 16 frames per second enven at a ridiculous small ROI of 904 x 904 pixels for H-alpha.

For white light I have a newer camera which makes about 82 frames persecond in 16 bit colour depth. Why do I not use that one for H-alpha. Simple, because the oldi siftware allows me to apply a false colour mask and so I can acquire Prominences and surface details in one single run. Unfortunately the author of th software Lucam Recorder is not interested anymore in adapting his software to more modern cameras like the ZWO and it is stuck to the brand Lumenera and The Imaging Source cameras. :cry: :cry: :cry:


regards Rainer

Observatorio Real de 14
San Luis Potosi Mexico

North 22° West 101°
Bruce G
Ohhhhhh My!
Ohhhhhh My!
Posts: 177
Joined: Mon Jul 16, 2018 8:18 pm
Has thanked: 128 times
Been thanked: 121 times

Re: Does this appear to be overprocessed?

Post by Bruce G »

Rainer,

I'm afraid I don't have my own Dropbox. It's $20 a month for files this large and I just don't send that many files

Sorry

Bruce G


Post Reply