Solar H-alpha close-up views - Forums

Topic starter

12 days ago

Hi all !

I’m relatively new to imaging and even newer to solar H-alpha observation and imaging. I don’t have much experience with other solar scope but I can tell the views on sunspots and prominences through my setup are gorgeous! I’m trying now to improve in solar imaging. I was so far using more or less the same acquisition and processing pipelines for solar as for planetary: 8bit acquisitions on ROI for super high FPS, stacking with AS! and sharpening with Registax. I’m now trying to switch to ImPPG which I have understood is very powerful and more adapted than Registax for solar H-alpha.

My images are always close-up views as the sweet spot of my H-alpha filter is small. So close-up view of regions of the disc (sunspots) or of prominences. Starting with Sunspots, would you tone-invert it or not? My issue is: tone inversion helps to bring this 3D effect and brings the fibrils up. But either we end up with white sunspots or with new bright regions close to the sunspots which are not naturally bright. In full-disc images, I feel like these are not so distubring, but for close-up views, I don’t like them.

Please see my example in https://app.astrobin.com/u/astro_yeye?i=g97wpl where you can find the post-stacking tiff files. They are based on an 8bit acquisition although I’ve just understood that 16bit is better for solar imaging (will try it next time).

This is the original stacked image:

Sharpening and leaving the tone curve as is, just a little stretch:

Bringing the high intensities down: bring some nice 3D effect

Inverted: white sunspots

Inverted by try to keep the sunspots black: bright unnatural ring around the spot

What are your thoughts and what is your experience with using ImPPG? What are your optimal settings or how to derive them?

I’m focusing on the tone curve, but if you have advice or feedback on anything else, you would be very welcome!

Additional comment/question is: zooming in ImPPG, I’ve noticed vertical and horizontal lines:

They don’t seem to appear in the final image. Is this just a display thing? Do you have already seen that?

Well written Respectful Engaging

Philippe Fossier

10 days ago

In my experience and being in contact with other solar imagers, Imppg is a standard tool used for sharpening and adjusting contrast but it is desirable to have more than one tool in your toolbox.

We also use Bill Blanchan's Solar Toolbox for PixInsight ( https://www.youtube.com/watch?v=3yzfH5x5Smg ).

It has a set of tools that are more targeted at bringing out proms, managing contrast and sharpening.

Imppg works well most of the time, but I often compare with the Solar Toolbox to see which wil give the best results.

Well written Helpful Respectful Concise Engaging Supportive

GreatAttractor

10 days ago

Hi Olivier,

Inverting or not is up to your personal taste (I myself switched to positive-only approach years ago). Your captures look good - though I’d decrease the sigma parameter a bit, it seems the fine linear details are a bit bloated/overcooked.

As for the graphical artifacts on zooming - I’ve seen it on one of my machines with outdated graphics drivers; updating them might help in your case. If not, you can switch to the CPU back end (menu option Settings/Processing back end).

Olivier Girard · Apr 20, 2026, 08:50 PM
They are based on an 8bit acquisition although I’ve just understood that 16bit is better for solar imaging (will try it next time).

Definitely, go ahead and experiment; but in my experience, it’s not really needed unless you’re trying to e.g. capture very dim prominences and stretch them heavily later. For what it’s worth, all solar pictures/animations currently in my gallery were captured in 8-bit mode.

Well written Helpful Respectful Concise Engaging Supportive

Philippe Fossier

10 days ago

Solar images #4 & #5 look good to me. The last one shows excessive R-L deconvolution. After a while, you will recognize the signs of excessive deconv.

On the question of 8 vs 16 bits - I agree. H-alpha is pretty much monochromatic to start with and 8 bits is sufficient (I can share some articles on the topic that expain why). At 8 bits you can capture at higher fps.

Keep in mind that solar imaging comes down to mostly 2 factors:

seeing quality
frame rate (higher is better)
Normally, exposure can be determined as: 1000/fps, meaning that if, for your ROI, your camera can give you 100fps, you can set your exposure at 10ms and lower your gain based on your histogram. Reality is that you should consider as low an exposure as you can, even if you need to increase your gain (within limits). This means that at 100fps, if you can capture with an exposure that is closer to 5-6ms, you will “freeze” the seeing and get better results.

Well written Helpful Respectful Engaging Supportive

Olivier Girard

Topic starter

8 days ago

Hi Philippe and Great,

Thanks for your feedback ! Your profiles are full of amazing pictures !

Philippe Fossier · Apr 22, 2026, 07:51 PM
We also use Bill Blanchan's Solar Toolbox for PixInsight ( https://www.youtube.com/watch?v=3yzfH5x5Smg ).
It has a set of tools that are more targeted at bringing out proms, managing contrast and sharpening.

I’ll check this out, thanks !

Philippe Fossier · Apr 23, 2026, 07:09 AM
Solar images #4 & #5 look good to me. The last one shows excessive R-L deconvolution. After a while, you will recognize the signs of excessive deconv.
GreatAttractor · Apr 22, 2026, 09:19 PM
Your captures look good - though I’d decrease the sigma parameter a bit, it seems the fine linear details are a bit bloated/overcooked.

Ok, I need to develop a form of feeling for this. I’ve found the doc here rather helpful: https://greatattractor.github.io/imppg/tutorial/tutorial_en.html

Writing this, I realise that ImPPG is your baby !?! ❤️❤️❤️

Philippe Fossier · Apr 23, 2026, 07:09 AM
Normally, exposure can be determined as: 1000/fps, meaning that if, for your ROI, your camera can give you 100fps, you can set your exposure at 10ms and lower your gain based on your histogram. Reality is that you should consider as low an exposure as you can, even if you need to increase your gain (within limits). This means that at 100fps, if you can capture with an exposure that is closer to 5-6ms, you will “freeze” the seeing and get better results.

I generally set the exposure time to 2-5ms and then play with the gain to “fill” the dynamic range. So the gain will be low for a close-up view somewhere on the disc and higher for a prominence. But for prominences, with 8bit, definitely the disc is saturated doing so. Can this mess up alignment in AS! ? Is it better to keep a reasonable gain and stretch (or adaptative-stretch) the prominence more heavily in processing? Couldn’t 16bit help here?

Respectful Supportive

GreatAttractor

8 days ago

Olivier Girard · Apr 24, 2026, 03:48 PM
Writing this, I realise that ImPPG is your baby !?! ❤️❤️❤️

Indeed.

If you’d like to send me raw stacks, I can have a look and suggest ImPPG processing settings.

Well written Respectful Concise Supportive

Olivier Girard

Topic starter

8 days ago

GreatAttractor · Apr 24, 2026, 04:53 PM
If you’d like to send me raw stacks, I can have a look and suggest ImPPG processing settings.

Nice, thanks for taking the time to help !

I’ve put the raw stacks (tif) as revision B of my image on my profile:
https://app.astrobin.com/u/astro_yeye?i=g97wpl&r=B#gallery

I have several other previous images, all sharpened with Registax and I intend to improve them with ImPPG when I get the chance. I would f.ex. be keen to have your feedback on how to stretch and sharpen close-up views of prominences, like this one (again raw stack in revisions B):
https://app.astrobin.com/u/astro_yeye?i=4udn3c&r=B

Thanks again !

Respectful

Philippe Fossier

8 days ago

do you have .tif files ?

.jpg are final images and difficult to work with

Olivier Girard

Topic starter

8 days ago

Right, sorry, I thought that adding my raw stacks tif as revision would make the job but it was jpeg compressed. I added the raw stacks as uncompressed source:

https://cdn.astrobin.com/uncompressed/336772/2026/40325fdf-e1dd-4a3f-93be-4960ae843e6b.tif

https://cdn.astrobin.com/uncompressed/336772/2026/571970e1-4817-4125-8a23-6f63c4e258b6.tif

GreatAttractor

7 days ago

Here’s my take on them (you can rename the settings files from .txt to .xml, but ImPPG can also load TXT directly).

📷 403.png403.txt

(Note that I lifted the tone curve a bit on the low end to highlight the penumbral details.)

📷 571.png571.txt

Olivier Girard

Topic starter

7 days ago

GreatAttractor · Apr 25, 2026, 07:49 PM
Here’s my take on them (you can rename the settings files from .txt to .xml, but ImPPG can also load TXT directly).

Woohaa, thanks a lot for taking the time and sharing the settings. I really like both. The processed sunspot is stunning !

I’ll take this settings and Philippe’s suggestion on using iteratively Gemini to help refine the settings as a basis to perform some trials on my past images.

Also, one of my initial interrogation mark was about image inversion. Leaving sharpening on the side, after quite some read and watching many tutorials, I strongly believe that it is a question of taste. Inversion can enhance the details and give a little more crisp, but by default, obviously, sunspots become bright. To me, I prefer to sacrifice a little crisp for a more physically meaningful result. There’s probably a way to keep spot dark while using inversion, with an “M” tone curve or something. I’ll continue to play.

Helpful Respectful Concise Engaging Supportive

Olivier Girard

Topic starter

6 days ago

I tuned a little bit the parameters for the prominence, in particular the low tones: I wanted to reveal the faint gas escaping horizontally. This has of course the cost of increasing the background level. I used an adaptative unsharpening mask to blur this background. It’s not perfect, but I’m quite happy with it!

https://app.astrobin.com/u/astro_yeye?i=4udn3c

imppg_setting_prominence_1_adaptative.txt

Well written Respectful Engaging

GreatAttractor

6 days ago

Looks very good. It’s often tricky with those surface+prom shots; I tend to either keep the disc normally exposed and sacrifice dimmer details in prominences, or just crank the exposure up and have full dynamic range on a prom and an overexposed disc. (That might change once I transition to a double-stack setup.)

Well written Engaging Supportive