Another solar imaging query

[lukebl]

Hi all. I have a query regarding the clarity and depth of detail in my Ha solar images.

With planetary imaging, the finished processed result after stacking, say, 3000 frames generally shows a lot more fine detail than can be picked out visually. However, with Ha solar imaging with my PST I have found the reverse to be true. The views through my TAL100RS / PST mod this morning were magnificent, with masses of crisp fine detail of the surface, filaments and prominences. Yet a 1000 frame stack shows a lot less detail, whether stacked in Registax or Autostakkert. Some individual frames even seem to show finer detail than the completed stack. Any ideas what I might be doing wrong?

[Merlin66]

I'd put it down to the seeing conditions...

The eye can focus/ refocus/ move around, to get the best "signal/data" to the brain...

The camera is "locked in place" and it's only the few fleeting moments of good in focus seeing which make it to the 20/1000 frames recorded being used.

Keep practising.......

[lukebl]

Thanks, Merlin. I totally get the importance of good seeing and practice. But the point I was making was that the same rules surely apply for planetary imaging, so why should the visual view appear sharper than a stacked image when the reverse seem true for planetary imaging? Well, to me anyway.

Anyway, I accept that I have a heck of a lot to learn about solar imaging!

[allcart]

How many frames did you take Luke. Why are you stacking 1000 frames.?

If you use AS!2.

Set image stabilization to surface,

Quality estimator to gradient, noise robust to 5 then click Analize.

When finished you will see your quality graph showing your stack of pics in order of goodness

Now.... at the top of the pic is a slider bar called frames, with a little pointer at the end. ; click this once and drag it along the bar. This will show you the frames in best order. You can also click once then use the arrow buttons on the keyboard to check the frames one at a time. You will see the quality degrade as you move along the bar.

Now......as you check the frames, you will see some info at the top corner of the image window.

There are 4 items.

1st line says mono or colour

2nd line tells you how many frames in the vid in brackets, but the last number with the # tag tells you what frame you are looking at.

3rd line tells you the quality of that frame in a % value.

What I do is........move the slider along the bar until the quality drops down to 80% . Now look at line 2....in brackets it tells you how many frames out of the stack you are looking at ( 56/1000)

Now if you choose 56 frames to stack, you will only stack 56 out of the 1000, but they will all be above 80% quality.

There ain't no point stacking bad quality frames if they just degrade the image. I have stacked only 7 frames from 1000 and still had a good pic.

I also find that placing alignment points manually is better than using multi points in a grid. I only place a max of 5 points on contrasty, clearly defined places on the best frame and this speeds up the process and does (in my opinion ) a better job.

Hope that helps.

[lukebl]

Very useful tutorial there, Allan. Many thanks. I have almost no experience of Autostakkert, so I really must find the time to get to grips with it.

[allcart]

So far I've found it to be the best for solar stacking. Its so intuitive and well designed. Reg works as well, but I find AS!2 so much quicker. You can use the saved images with wavelets in Reg5 or 6 if you understand wavelets. I can't get on with them, so I process in CS5.

[lukebl]

I've had a bit of a play with Autostakkert, and stacking just 200 of 1000 frames, but I'm getting the same results.

This picture shows the issues. The one on the left is the stacked result, tweaked a bit in Photoshop and Registax. The one on the right is a single frame. As you can see, the single frame, apart from a bit of noise, seems to be more detailed that the stacked result. I would therefore have assumed that the stacked result should be even sharper still?

[allcart]

I would therefore have assumed that the stacked result should be even sharper still?

Not necessarily, where proms are conscerned.

You have to bear in mind that the prom is a fast moving , dynamic, evolving entity. Unlike static objects like the moon, No two frames taken 1 second apart will be the same. For proms I would suggest short 20-30 sec vids., and stack only the top percentage. The more frames in a prom stack, the less defined it may become.

You can also up the sharpening a bit in photoshop to bring out the prom. There is a lot of experimenting to do to get the best you can from an image. I haven't mastered it yet, but I keep getting closer each time.

Try stacking the top 100 frames, then open in PS. Now apply an unsharpmask at 500%, radius 1.5 and see what happens. Hopefully the prom will sharpen up a lot. Maybe even too much, so you do it again with less. then a touch of Guassian blur, level, unsharpmask at a much lower radius, curves etc etc.

[allcart]

I had a little tweek just to try and show you what I mean. Bear in mind that I am dealing with a tiny image. I just masked out the sun and applied about 0.6 radius usm to the prom.

[Freddie]

I think the key is going to be your wavelet work. This is a pre wavelets stack

and this after wavelets

I use linked with small adjustments to 1 and 3 and a bit more on 2. Hope that helps.

[lukebl]

Thanks for the input, guys. All this learning is great for keeping my brain working! And there seem to be a lot more people on the forum doing solar imaging than when I joined a few years ago.

[allcart]

I found the learning curve was greater with dark side, so I joined the light side.

When I go back to dso's in the winter, I'll be asking you for advice.