NGC6826 another question...

[mcbbcn]

Hi,

I've been working on my NGC6826. I'm working with fairly small amount signal to figure out the flow/steps a bit faster -as I do quite bit of trial and error-. My first question is, does StarTools want to work with the .FIT file that comes out of my Pixinsight calibration process, or do I need to save it as a .TIFF? I don't know why but ST seems to render the initial 'autodev' color gradients differently if it's a .TIFF or a .FIT.

I thought I would attach the a couple of quick snips. One snip (the one that you can see smooth with no noise), it's just a simple Pixinsight screen stretch after I've done the calibration, alignment, and integration. The other one (the noisy one) is once I take the file from Pixinsight into StarTools, I just do a simple 'autodev' and wipe for gradients, and that is what I see...So, my second question is why is StarTools stretching the whole thing so much that you can even see the faint doughnuts left by the flat subtraction?

Also, I attach the links to the original .FIT and .TIF, in case you want to inspect the image...but I guess my 3rd question is, how come I can achieve such a quick balance in Pixinsight with a simple screen stretch but once I take the image into StarTools the stretching is so huge that I don't know how to move forward. So, if somebody would like to use my links to the original files and give it a go, and see if you can achieve a good balance in StarTools, I would love to know your flow, so I can learn it too.

https://dl.dropboxusercontent.com/u/897 ... n_GBRG.fit

https://dl.dropboxusercontent.com/u/897 ... n_GBRG.tif

Thanks for your help,

M.

[admin]

Hi,

I've been working on my NGC6826. I'm working with fairly small amount signal to figure out the flow/steps a bit faster -as I do quite bit of trial and error-. My first question is, does StarTools want to work with the .FIT file that comes out of my Pixinsight calibration process, or do I need to save it as a .TIFF? I don't know why but ST seems to render the initial 'autodev' color gradients differently if it's a .TIFF or a .FIT.

There should be little to no difference between TIFF and FITS file. FITS is preferred though, as it is as close to the raw signal as possible

I thought I would attach the a couple of quick snips. One snip (the one that you can see smooth with no noise), it's just a simple Pixinsight screen stretch after I've done the calibration, alignment, and integration. The other one (the noisy one) is once I take the file from Pixinsight into StarTools, I just do a simple 'autodev' and wipe for gradients, and that is what I see...So, my second question is why is StarTools stretching the whole thing so much that you can even see the faint doughnuts left by the flat subtraction?

"It's not a bug - it's a feature!"
AutoDev was designed to find the absolute optimal stretch for bringing out all detail in your image. As you can see it is really very good at that. Unfortunately it's detail that you would like to bury in this case, but you can't fault AutoDev for bringing out the donuts; as far as AutoDev is concerned it's 'interesting' detail. AutoDev is indeed very good at showing the paint points as well in your data.That's why it's the first thing I do when I get new data; AutoDev. It'll show gradient, stacking artifacts, dead pixels, hot pixels, noise, dust donuts, etc. etc. I then fix the issues to the best of my ability and often go back to AutoDev for my final stretch. You can tell AutoDev to only optimize a certain area (by specifying a Region Of Interest), and you can tell it to ignore the finer detail such as noise.

Like this;

Screenshot-Untitled Window-1.jpg (175.91 KiB) Viewed 8101 times

(notice the green ROI rectangle over NGC6826)

AutoDev will now only optimize the global stretch for that area.

Add some deconvolution and color calibration to the mix (the latter being a little harder because of the LP filter used) and you end up with something like this;

NGC6826_integration_GBRG.jpg (12.72 KiB) Viewed 8101 times

Full log;

--- Auto Develop
Default settings
--- Crop
Parameter [X1] set to [1440 pixels]
Parameter [Y1] set to [904 pixels]
Parameter [X2] set to [1872 pixels (-1456)]
Parameter [Y2] set to [1174 pixels (-856)]
--- Wipe
Default settings.
--- Auto Develop
ROI over NGC6826.
Parameter [Ignore Fine Detail <] set to [3.6 pixels]
Parameter [Outside ROI Influence] set to [1 %]
--- Deconvolution
Modifed autogenerated mask a little to include more of NGC6826 (be careful not to cause ringing).
Parameter [Radius] set to [3.9 pixels]
Parameter [Iterations] set to [6]
--- Color
Parameter [Cap Green] set to [To Yellow]
Parameter [Style] set to [Artistic, Detail Aware]
Parameter [Dark Saturation] set to [4.70]
Parameter [Bright Saturation] set to [Full]
Parameter [Saturation Amount] set to [278 %]
Parameter [Green Bias Reduce] set to [1.60]
Parameter [Red Bias Reduce] set to [1.55]
--- Wavelet De-Noise
Parameter [Brightness Detail Loss] set to [16 %]
Parameter [Redistribution Kernel] set to [4.5 pixels]
Parameter [Read Noise Compensation] set to [Off]
Parameter [Smoothness] set to [75 %]

And that's it!

[mcbbcn]

Wow! You're amazing! And the process feels a bit like magic -good magic-. I'll study your steps and I'll work to improve my ST technique. I really appreciate the detail explanation.

M.

[mcbbcn]

I processed the Image having the 4.3 hours of information and adding Ivo's steps plus a few other things he's taught before, and this is the result...remember that it's is a Bortle 8 sky...I'm basically in Portland Downtown. Cheers, M.

[admin]

That's just excellent. I got nothing to add!

[mcbbcn]

That's just excellent. I got nothing to add!

Thanks a lot Ivo! I could have not made it without your teaching.