15.4.2020 Hedgerow prom at 8 o'clock on SE limb..

[Freddie]

Chicken and bacon pie with assorted veg. Very nice!!

The difference in detail (particularly between the 10 frame and 100 frame) is quite hard to see because of the noise and is only really easier to see in the brighter parts of the image.

A couple of things I would change in your capture to make processing easier would be to really try and get your gain down. I see you have reduced it to 262 but I would really get this well below 200. Raising the exposure a couple of ms really won’t make any difference which will allow you to reduce gain to achieve the same level of histo but I would reduce further as the edge of the disc is over exposed which is burning out the small spicules. Reducing the overall brightness and increasing the exposure a few ms should give you plenty of scope to reduce the gain quite a bit.

Stacking 10 frames was good to show the difference for the experiment but I think stacking 50 would be more realistic. I always think that setting a number rather than a % is a much better approach to stacking. 

What has happened in the post processing in imppg is that the sharpening has been applied globally. At least that’s what it looks like to me. You will have to correct me if I am wrong. What has therefore happened is that it has sharpened the proms but as there is plenty of bright signal there, the sharpening has not shown up the noise so much. However as the sharpening has been applied globally (I assume) the same amount of sharpening has been applied to the background and as this has very little signal, it has pretty much sharpened the noise, resulting in that bright mottled pattern. The noise in the larger stacks has had the same sharpening applied (as you say you did the same to all 3 images) but as there are more frames in the stack, the noise is smoother and therefore doesn’t result in that mottled effect. You are right that the background of the 100 and 1000 stacks are much superior so what we need to achieve is the same background as the 100/1000 stacks but with the improved detail (to my eye at least) of the smaller stack.

The way to do this is to sharpen the 10 stack (or something like 50 if it wasn’t an experiment) in imppg but use the “adaptive” settings. This allows you to sharpen brighter and darker parts of the image to different degrees. In fact you can actually apply smoothing rather than sharpening to darker parts of the image with the right settings. This way you keep the detail in the bright bits and don’t sharpen the noise in the dark areas.

Hope that helps.

 

 

[Rusted]

Thank Freddie.

I haven't really used the adaptive settings in ImPPG. I will certainly have to give them a try.

Until now I was never sure at which point the exposure time limits the thermal agitation I see on the monitor.
My concentration on ultra-short exposures has been at the expense of increased gain.
You and Merlin have both given me a number of things to try. We shall see which approach works for me.
Once a video is captured it can be manipulated in a number of different ways.
But now it is time for my dinner. Thank you both for your input. 

[Rusted]

Continuing on the topic of number of stacked frames:

First I captured a 2000 frame video of today's SE prom in rather poor seeing conditions:
I then ran it repeatedly through AS!3 [Autostakkert] using increasing numbers of stacked frames.
The results were then passed [identically] through ImPPG and finally cropped in PhotoFiltre and assembled into one image

To my eyes the 100 from 2000 frames offers the best prom detail:
More frames becomes progressively softer. Fewer frames exaggerates noise.
I gave each image a file name showing the number of stacked frames. 10fr, 100fr, 1000fr, etc.

 

[Freddie]

As suggested, the more frames in the stack the more poor quality ones in there which will blur the detail. I think we can ignore the 10 stack as that isn’t realistic for a final image (fine though for an experiment) but I think the 50 stack would be the best with a few adjustments during capture and processing as I think the noise in the 50 stack is masking detail.

Still looks like you have quite a high gain setting and the overall exposure is still quite high. I would reduce gain further until you get a slight darkening of the edge of the disc. If at that point gain is anywhere close to getting up to 200, I would increase exposure by a couple of ms until gain is well below 200. Then in processing, do try the adaptive settings to stop sharpening the background.

With these changes, I think you will further improve on your images.

[Rusted]

Thanks Freddie.
I preferred the 100 frame image on the full image size but it was a difficult choice.

I am really struggling to get exposure and Gain low with my 2x Barlow and 640x480 frame size.
I'll get rid of the Barlow and try a larger frame size. The image is boiling madly at this scale.

[Freddie]

I agree that of the shots posted, the 100 stack is the best. I however think the 50 stack approach has more potential with the changes I have suggested.

You need to keep the Barlow in to achieve the image scale required so I would leave that. A larger frame size will make no difference when it comes to gain it will just slow down the FPS you can achieve as it is having to download more data.

I still think you have room to lower gain as you look a touch over exposed so lowering gain will sort that and a slight increase in exposure by a couple of ms to allow a further lowering of gain will generate more benefits from reduced noise than you will loose by having exposure a few ms longer. Anything even up to 10ms really won’t cause you any problems.

Once you are getting the best from captures and stacking, you can then concentrate on improving post processing with careful sharpening but only where you have bright signal that can cope with it and not the dark low signal areas.

[Merlin66]

For proms...based on some animation trials I'm doing, I think 2000 frames may be working against you.

AS3! sorts the "best" frames but these are selected from the whole pile of 2000 frames....if there's movement within the prom in the time taken for the video (50 sec or so?) then stacking the "best" may give a poor result.

I'm thinking of a run of 400-500 frames (10 sec or so) and going for the top 100 frames may be a better strategy.

Just my 2c

 

[Rusted]

Thank you both for all your helpful advice. I'm sure it will also be useful to a much wider audience.
You have given me plenty of different things to try now and to think harder about what I am doing.

My previous imaging routines had become rather inflexible. Usually 40% of 3k and lots of wishful thinking.
Suitable only for perfect seeing conditions with my 150mm/ 6" aperture.

Your expertise clearly shows that I was making far too many false assumptions.
I was convinced that 1-2ms was the only way to freeze thermal, image movement.
That huge video files were a good thing. Now you talk of 10ms and much smaller numbers of frames.

Tomorrow offers yet more sunshine for me to practice what you preach. 
Though it may be slightly premature to suggest that you will be astonished. 

Thanks again.

[Freddie]

No problem, glad it was of some help.

I'm assuming that your 2000 frame captures are only taking about 10 seconds so the movement of the proms over time will not be an issue over that timescale and image scale.  Merlin 66, the numbers you are quoting would indicate you are only getting ~40 fps, that sounds slow if you are using your ASI.

Let us know how you get on with the adaptive settings in imppg.

Edited April 17, 2020 by Freddie

[Merlin66]

Freddie,

With the ASI 174 and a selected ROI I can get up to 400 fps on the spectroheliograph.

Imaging proms with the double stack SM60 on an ED80, sometimes I'm lucky to get above 15 fps with the ASI 1600 and ROI. That's also using gain 139 and gamma 50 under Firecapture.

 

[Freddie]

Makes sense you are getting the fps rate you are with that setup. Sounds like you are exposure duration limited rather than camera speed limited then.

[Merlin66]

I suppose you're correct.

Nothing that more aperture and larger ($$$$$) filters couldn't fix.

Edit: With the single stack/ED80/ ASI 1600 I can get around 50 fps for the proms.....

Edited April 18, 2020 by Merlin66

[Rusted]

I'm back.
Started at 8.00 capturing lots of short [mostly] 500 frame videos and a few 1k..
Stacking only 100 in AS!3. Then ImPPG.
Between 8am and 9am the seeing wasn't great.
It improved after 9.00 as the sun gained altitude.
Using 800x600 frame size in SharpCap with 2x WO Barlow on the ASI174.
Had to change to the PST BF for increased prom contrast over the Lunt B2000S2.
The PST BF allows much lower gain and shorter exposures.
The seeing went bad again after 10.00

Not proud, but the best I can do this morning.

 

[Merlin66]

Russ,

A pretty good result never the less...

We've been under cloud for the past few days.....

 

[Pete Presland]

Fantastic looking prom group