StarTools

Ivo,

I continue to learn how to make the most of Star Tools and I am finding that on some images ST blows away PixInsight and/or Photoshop and on other images, I can't seem to get what I want out of ST. I decided the best way to illustrate the point was to provide you a detailed example. I see that I am limited to 3 attachments per message so this will be continued on a reply.

A few night ago I got an hour of imaging on NGC6888, Crescent Nebula. I am now using flats and bias frames and my data is exponentially easier to work with. Capture details:

C8 XLT
F/6.3 reducer
Guided and dithered every 2 images
20 subs @180s each, with twilight flats, darks and bias

Here is the image after an initial Auto Develop - Default settings
I did a fairly tight crop to 3260x2064 to isolate the nebula and get rid of image issues at the bottom left and middle left. I didn't see any dark spots/dust motes so I did a default wipe with a full mask.
I then did another Auto Develop - Default settings

I then did a wipe with the default vignetting settings and a manual develop with digital development set at 90%. The Auto Dev was blowing out the nebulosity so I did it manually.
I saw a small blemish on the left edge so I lassoed it out of the mask and did a default gradient wipe.
I then did a default bin.

After the bin I used the Life Module to bring out the nebula. I used the Isolate preset with a full mask and 10 pixel mask fuzz.
Here is a closeup after the Life Module.
I then used Contrast, Decon, and Sharpen, all with default settings.

I then use the Life Module again (a mistake maybe??) Again I used the the Isolate preset.
I then opened the Color module (the new color module is really nice!) This is the image after opening and before any tweaks.
I capped green to yellow, set saturation to 300% and dark saturation to 2.5

I then stopped tracking and launched the denoise module.
Finally I used the Filter module to try to increase the fainter nebulosity.
The image still needs some work (I overdid the Filter nudges) but as is it illustrates the issue I continue to have with some images. What seems to be happening is that somewhere along the way, each star is increasing the brightness of the background sky around it out to some number of pixels. Where the stars are dense it creates a relatively uniform background that is lighter than deep sky black. Where the stars aren't quite as dense areas of deep sky black form what appears to be black channels through the areas of lighter areas around the stars. Where stars are sparse, such as int he top left corner, there are larger areas of back deep sky. I find the effect to be distracting.

I understand that stars will have halos, but beyond the halo I prefer black instead of the patches of lightness. The image of M51 that I posted a couple of weeks ago is a great example. The sky around M51 is relatively free of stars and the sky comes out quite dark.

I think the effect is being induced starting with the Life module and cemented in place when tracking is stopped. I ran Life twice to bring out the nebulosity but I think even if I used it just once I'd still see the problem. I would really love to figure this out.

The final image below is what I came up with using PixInsight. I also did a saturation boost in PhotoShop CC and did a couple of curves adjustments to bring out the ripples in the middle of the nebula more. Looking closely I can see the same black channeling in the PI image, but it's much fainter and less noticeable. I know I should be able to do as good or better in Star Tools but I'm just not there yet. If you'd like to see the data it's available at https://www.dropbox.com/s/9dvzqutm0zxft ... 20Subs.fit

A high res view of the PI/PhotoShop image is at http://www.astrobin.com/full/52108/?mod=none

Just to be clear Ivo, I am in no way complaining about ST. I love it and am continuously amazed at the power of its features. In this case I thought it would be more useful to illustrate my question and post the PI image to show what I know is in the data.

Thanks for posting this Gary.
First off, let me say that this is extremely helpful stuff you're posting here. It shows me where any possible painpoints are when comparing to other software and it certainly doesn't come across as complaining!

AutoDev is indeed a good starting point to examine the data, which you indeed do.
If you decide you'd like to bin your data, then do it as early on as possible. The reduced noise will help a number of modules achieve better results (chiefly the decon module).
Wipe tends to do a good job unaided, but when large parts of the image potentially have faint DSOs in them, it is wise to mask out the parts of the sky where you think they are. In this case, using the lassoo tool to fully mask out the larger structure (e.g. the whole of where you expect NGC6888 to be) will help Wipe target only the backround and will help it leave any nebulosity in.
In PixInsight terms, Wipe without a mask works like (a better) ABE. Wipe with a mask works like DBE. See also this video.

After performing the Wipe, you can do a manual Develop, or AutoDev with a Region of Interest. Both should reveal the full extent of NGC6888 now (however faint!).

Doing Decon as early as possible is also good practice, as some modules can then take into account deconvolved data, yielding slightly (slightly!) better results.

The Life module's purpose is to manipulate (very) large scale structures. If you are absolutely certain the larger structure is in fact in the data (and we have just proven that it is), then you can create a mask that outlines the larger structure that you are seeing. Life will then reconstruct the larger structure using hints of any faint nebulosity within the designated area, re-embedding the detail within it (the Isolate preset).

Do all this and you'll end up with something like this;
Of course, tweak to your personal taste...

Hope this helps Gary - do let us know how you get on!

Thanks Ivo!

I just saw the new video you posted on YouTube and now I'm itching to take another crack at my NGC6888 data. I'll post what I get.

Ivo,

I reworked my Crescent Nebula with what I picked up in your response above and your M8 tutorial. This is a huge improvement for me I think. There is a lot more nebula detail in this image than what I got out of PI.

Next I have to figure out how to prevent the inappropriately colored star halos. I knocked a lot of it down with the Filter Module. Ideally though I'd prevent it from happening rather than trying to fix it after it occurs.