Zerene Stacker now has dust & hot-pixel removal

Smokedaddy · Post by **Smokedaddy** » Sat Mar 28, 2020 3:15 pm

Rik,

As I mentioned before, excellent idea, thanks very much. Since you're a math guru are you familiar with DefectMap/Pixel Math?

-JW:

rjlittlefield · Post by **rjlittlefield** » Sat Mar 28, 2020 3:24 pm

Since you're a math guru are you familiar with DefectMap/Pixel Math?

Are you talking about https://pixinsight.com/doc/tools/Defect ... ctMap.html ?

If not, then can you provide a link or other reference?

--Rik

Smokedaddy · Post by **Smokedaddy** » Sat Mar 28, 2020 3:27 pm

... yep and Pixel Math.

https://www.pixinsight.com/doc/legacy/L ... _math.html

-JW:

rjlittlefield · Post by **rjlittlefield** » Sat Mar 28, 2020 3:45 pm

OK, yes, familiar with those. Is there some particular part that you wanted me to notice or discuss, or just making sure I knew the reference?

--Rik

Smokedaddy · Post by **Smokedaddy** » Sat Mar 28, 2020 4:13 pm

Nope, a couple of my Astro-buddies use these for masking. Just didn't know if you were aware of them. Feel free to delete my messages.

JW · Post by JW » Sun Mar 29, 2020 11:29 am

Thanks Rik, since we do Defense contracting our Admins have upped their game. Time to get my own PC, I guess

TheLostVertex · Post by **TheLostVertex** » Sun Mar 29, 2020 3:22 pm

I am planning on experimenting with this tonight, but I have an initial question. How does zerene handle the mask compared to other frames that are getting aligned? Is the mask aligned to the first image, or last image, or something else?

Also, it can be pretty hard to see dust spots in out of focus areas of an image, but once it is in focus the spot becomes very visible. Do you have any ideas on a good way to make the map for stacks that have hundreds of images, and with respect to the question about alignment?

Lou Jost · Post by **Lou Jost** » Sun Mar 29, 2020 5:00 pm

TheLostVortex, Rik will confirm, but I'm sure the mask must be applied to each frame before alignment.

rjlittlefield · Post by **rjlittlefield** » Sun Mar 29, 2020 5:54 pm

Yes, the infill operation is applied to images before alignment.

it can be pretty hard to see dust spots in out of focus areas of an image, but once it is in focus the spot becomes very visible.

Hhmm... In my experience it's the reverse: isolated spots are easier to see in out-of-focus images and become difficult to see when they overlay focused detail.

What I would do with a deep stack is layer up two or three different images in Photoshop, say one from the front, one from the back, and one from the middle, slap a layer over the whole set to become the mask, then toggle visibility on the lower frames so as to flash between them. Any spot that looks the same in two very different images is a candidate for masking.

One troublesome case this does not cover is where a spot is hard to see in a single frame, but makes a trail that is painfully obvious.

In the long term, my plan for those is to extend the retouching tool in Zerene Stacker so that it can retouch the mask also.

But in the short term this problem will require multiple passes, tweaking up the mask in Photoshop or whatever. In my first example, the weathered table, I found all the dust spots on the first run, but then it turned out that I had not thought to look for hot pixels, and there was one of those that made a trail of little bright dots. To fix that required an edit on the mask. If you look close at the mask, you'll recognize the hot pixel dot because it's smaller (just below left of center frame in the crop).

The "standard solution" for dust spots is to shoot a blank frame, light gray, right after shooting the real stack. Then levels-adjusting the blank frame can make the dust spots pretty obvious.

If you anticipate that multiple passes may be required, then consider removing the checkmark on "In-fill all frames" and instead put a checkmark on "Explicitly propagate good pixels". That's a lot faster, and while it usually generates an inferior stacked result that still has some spots, it generally kills trails. Any remaining trail then will belong to a dust spot or hot pixel that needs to be added to the mask.

--Rik

TheLostVertex · Post by **TheLostVertex** » Sun Mar 29, 2020 6:42 pm

rjlittlefield wrote:
it can be pretty hard to see dust spots in out of focus areas of an image, but once it is in focus the spot becomes very visible.
Hhmm... In my experience it's the reverse: isolated spots are easier to see in out-of-focus images and become difficult to see when they overlay focused detail.

To be clear about this, I have some stacks where the out of focus area is very "busy" with lots of color and detail across the whole frame, so the dust spots end up blending in with everything that is out of focus. I cant tell what isnt suppose to be there until it comes into focus. For other images with smoother backgrounds I dont have this problem. Until now I have tried using spot removal and clone tools in Capture One/Lightroom and copying and pasting between images with mixed results. Sometimes it works great, other times it takes so much tweaking on an image by image basis its not worth the effort.

I notice that you can have both "In-fill all frames" and "Explicitly propagate good pixels" checked at the same time. What happens then? I am assuming "in-fill all frames" only is the most desirable setting for image quality?

-Steve

rjlittlefield · Post by **rjlittlefield** » Sun Mar 29, 2020 7:00 pm

TheLostVertex wrote:I notice that you can have both "In-fill all frames" and "Explicitly propagate good pixels" checked at the same time. What happens then? I am assuming "in-fill all frames" only is the most desirable setting for image quality?

Your assumption is correct: In-fill all frames for highest quality.

At this time I don't know any circumstances where it would be a good idea to check both options. If you do check them both, then you'll get the high cost of in-filling, plus a high likelihood that explicit propagation will introduce spots that you would not have had with just infilling. I've left it possible to check both for now, because maybe some good use will turn up. But it's pretty likely that I'll decide to change that bit of the UI so that only one can be checked at a time.

--Rik

Luisifer · Post by **Luisifer** » Mon Mar 30, 2020 4:59 am

Superfunction.

It is pleasure to see that function in ZS. Because of i have action(s) in PS that retouche dust.... It would be ok but PS doesn't accelerate it so it runs very slow (for example one frame per two minutes). So in case of thousands frames it is neverending batch...

TheLostVertex wrote:it can be pretty hard to see dust spots in out of focus areas of an image, but once it is in focus the spot becomes very visible. Do you have any ideas on a good way to make the map for stacks that have hundreds of images, and with respect to the question about alignment?

As Rik mentioned, good choice is to start with frame where is the smallest part in focus. IMO is good to choose one of the last frame from all. Because the last frame should be the most dusted (and hot/dead pixeled) thanks to many exposures.

For those who are not familiar with "popping up" of the dust on photo, i use zigzag curve. It helps near to perfectly.

https://12in.cz/temp/dust.jpg

___
But antidust function disappointed me after the first test. Looks like it will be necessary to stay on slow PS batch.

rjlittlefield · Post by **rjlittlefield** » Mon Mar 30, 2020 9:20 am

Luisifer wrote:But antidust function disappointed me after the first test. Looks like it will be necessary to stay on slow PS batch.

I am guessing you mean that the Zerene Stacker anti-dust function disappointed you.

I notice that your dust spots are quite large, easily visible even as a single dot at web size. So, I am thinking that you prefer to have the spot treated with something like Photoshop's "content aware fill" rather than Zerene Stacker's smooth gradients.

Is that correct? If not, can you explain further?

Also, can you explain further how your Photoshop action works? I'm sure that other people would like to use that approach also.

--Rik

Luisifer · Post by **Luisifer** » Mon Mar 30, 2020 9:46 am

Yes, you are right. I will make few tests for comparsion.

I prepare action in photoshop where i use spot healing brush tool. When i record action it is set to record brush strokes (this part is not accelerated - so this is reason why it is so slow). So i record action on "zigzaged" frame. And after that i use this action on each frame from stack (PS - File - Automate - Batch).

mawyatt · Post by **mawyatt** » Mon Mar 30, 2020 10:20 am

Isn't the "zigzagged" caused by a hot pixel that is fixed in the sensor?

If, so then each image would not have a zigzag but only a a hot pixel at a fixed location in the sensor. The zigzag is the result of stacking, thus zigzag correction should NOT be applied to each image, only to the image rendering of the stacking.

Best,