Ca-K Processing Experiment

[David Smith]

Had some time on my hands this morning so spent it productively wandering SGL Came across this thread on ImPPG workflow

Decided to have a play around applying this to my Ca-K FD taken on the 9th. My processing has always been pretty basic and whenever I get the urge to try and improve it I end up having to do something else, like go to work or I've been efficient and deleted all my raw data! This time however that was not the case and I actually managed to spend enough time to get a result that I think is an improvement not just different! So first up is my FD from the 9th:

It's far from my best work, a little bright in the centre and the panes are all affected slightly differently by variable seeing. Not much I can do about that but going back to the stacked frames that came out of AS!3 and using this workflow as a guide I ended up with this

Hey presto I have proms in what I thought was just a surface shot! Still needs some work:-

1. I need to leave more space around the limb in each frame in future so as to allow room for any proms

2. I still don't understand how to use adaptive sharpening in ImPPG effectively. If anyone knows of a good tutorial on this I would be most grateful.

Any thoughts, comments or suggestions as to what else I might try would be appreciated.

Strange how things play out, I almost imaged yesterday but found the seeing was a tragedy and decided not to bother. Had I done so I would have cleared my hard drive to maximize available space and this may never have happened!

Forecast is for clear skies tomorrow,suffice to say I will be eager for some more data to play with and probably disappointed too knowing UK weather.

Thanks @Nigella Bryant

[Altocumulus]

Try this one - I'm still struggling a bit with the tone curve, especially the way this tutorial suggests moving points...

 

 

[Pete Presland]

Interesting difference between the images, I also am not sure about adaptive sharpening.

[michael.h.f.wilkinson]

4 minutes ago, Pete Presland said:

Interesting difference between the images, I also am not sure about adaptive sharpening.

Adaptive sharpening is an idea I discussed with Filip (the author of ImPPG) some time back. I implemented an adaptive unsharp masking in C, which varies the strength of the unsharp masking as a function of local brightness, based on the idea that bright areas have better S/N, so can be tackled more aggressively, whereas darker areas need a gentler approach. The principle works, but it is difficult to set the parameters right, I find.

[Freddie]

Nice looking images.

There have been a couple of questions about tone curve/histo and adaptive in ImPPG so I will try and help and build on what Michael has said.

Adaptive: This basically allows you to control the amount of sharpening applied to dark (low signal) and bright (high signal) parts of the image. This is because if you apply the same amount of sharpening across an image, the sharpening will be fine in the bright parts but will introduce noise to the darker parts or will be fine in the darker parts but the brighter parts would take more sharpening. ImPPG controls this by allowing you to adjust “amount min” the amount of sharpening in the dark parts and “amount max” the amount of sharpening in the bright parts. The thing that defines what is dark and what is bright is “threshold” so adjusting “threshold” will change the brightness level where the transition from amount min and max occurs. The “transition” setting basically allows you to set a band of brightness over which the change from amount min and max occur so it isn’t a sudden change.

The best way to get into this is probably fire up an image, apply a lot of sharpening and set amount max really high ~15 (a lot of sharpening) and min to say 0.6 (a bit of blurring) Now just look at the dark sky background of the image and adjust the “threshold” and keep looking at the dark part of the image, as you adjust it you will see the background become noisier as sharpening is applied to the dark region, now just adjust the other way and eventually you will see the background go nice and noise free as the background changes from being sharpened to blurred. Adaptive!!!! Interestingly the person in the video has this exact problem as he didn’t like the sharpening as he said it added to much noise to a dark part by the prom so he dialled back the sharpening a touch. He made no mention of adaptive and didn’t have it set so I guess he wasn’t aware of how to use it.

Tone curve/histo: Best thing to do is ignore moving points up, down, left or right and just think about what the point/curve actually means. An unchanged image will show pure black on the left and pure white on the right and all shades of grey in between. The tone curve is the line between all these points. The brightness of any point on the curve can be adjusted by moving the curve from its original diagonal position. Obviously pure black can’t be made any darker so the far left can’t be lowered and pure white can’t be brightened so can’t be raised. Any point in between can however be adjusted. Again best way to do this is by looking at an image. Look at the dark background, put a point on the left part of the curve, move it up and the dark part gets brighter but notice, it’s not just the very dark point you picked that has got brighter the dark grey parts of the image have also become brighter. Look at the curve though and you will see that the by dragging up a single point has created a nice curve in what was once a straight line so any parts that are above where they were on that original line will have also become brighter, not just the point you dragged up.

Now look at why you sometime get some odd things happening when you have a couple of points selected on the curve. Fire up an image again and start with an unchanged tone curve i.e. a nice diagonal line. Just put a point in the middle of the curve don’t change the shape of the curve, leave it as a straight diagonal line. Now go to the left side of the curve and as you did before, drag a point up to brighten the dark parts of the image. The dark parts brighten but strange things also happen to the lighter parts. Look at the curve, the point you put in half way along has acted like a pivot and the curve to the right of it has now gone down and we know from before that as the curve there is now below it’s starting position those points have now become darker. This is why you need to be careful when you have multiple points on a curve as an adjustment to one point on a curve can have unexpected consequences to another part of the curve unless you keep an eye on how that curve is changing. If a part of the curve has moved that you didn’t want, just drag it back into place or put a point on the curve which will prevent that particular point from moving.

Hope all of that is of some help.

 

[Nigella Bryant]

6 hours ago, David Smith said:

Had some time on my hands this morning so spent it productively wandering SGL Came across this thread on ImPPG workflow

Decided to have a play around applying this to my Ca-K FD taken on the 9th. My processing has always been pretty basic and whenever I get the urge to try and improve it I end up having to do something else, like go to work or I've been efficient and deleted all my raw data! This time however that was not the case and I actually managed to spend enough time to get a result that I think is an improvement not just different! So first up is my FD from the 9th:

It's far from my best work, a little bright in the centre and the panes are all affected slightly differently by variable seeing. Not much I can do about that but going back to the stacked frames that came out of AS!3 and using this workflow as a guide I ended up with this

Hey presto I have proms in what I thought was just a surface shot! Still needs some work:-

1. I need to leave more space around the limb in each frame in future so as to allow room for any proms

2. I still don't understand how to use adaptive sharpening in ImPPG effectively. If anyone knows of a good tutorial on this I would be most grateful.

Any thoughts, comments or suggestions as to what else I might try would be appreciated.

Strange how things play out, I almost imaged yesterday but found the seeing was a tragedy and decided not to bother. Had I done so I would have cleared my hard drive to maximize available space and this may never have happened!

Forecast is for clear skies tomorrow,suffice to say I will be eager for some more data to play with and probably disappointed too knowing UK weather.

Thanks @Nigella Bryant

No probs, brilliant. You have to disable the Lucy deconvolution by clicking on the box and then adjust sigma, not much, then I increase the max slider,dont normally touch much else, don't overdo the sharpening, forget the name of that as I'm not at my computer.

[NiCo10]

These images look fantatsic!