StarTools

Ivo, I remember you working on this previously, but about half the time I try the decon module I still get some random white pixels created (fewer than previous versions, but they are not gone). Playing with the radius and perhaps moving the other slider to "noisier" might fix it, but usually not where I'd wanted those sliders to end up.
I don't know if you are aware of this and still looking at it, or gave up etc, so if you want another take on it and you can't easily recreate it I can provide sample data.

Argh. Not again!

I thought I had squashed that bug.

Any pointers on how to replicate would be greatly appreciated, thanks!

Hmm, I tried cropping the image to an area where this appears, and process that, but it doesn't happen. So I guess you'll have to go with my full image: https://drive.google.com/open?id=0B1c3j ... nhqLXBmYzQ

Simple steps:
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside ROI Influence] set to [15 %]
--- Bin
Parameter [Scale] set to [(scale/noise reduction 50.00%)/(400.00%)/(+2.00 bits)]
--- Crop
Parameter [X1] set to [32 pixels]
Parameter [Y1] set to [32 pixels]
Parameter [X2] set to [2569 pixels (-32)]
Parameter [Y2] set to [1700 pixels (-32)]
--- Wipe
Parameter [Mode] set to [Correct Color & Brightness]
Parameter [UNKNOWN] set to [No]
Parameter [Precision] set to [256 x 256 pixels]
Parameter [Dark Anomaly Filter] set to [4 pixels]
Parameter [Drop Off Point] set to [0 %]
Parameter [Corner Aggressiveness] set to [100 %]
Parameter [Aggressiveness] set to [75 %]
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside ROI Influence] set to [15 %]

(auto develop without ROI - hey perhaps you should store the ROI rectangle in the log for cases where it might matter!)

And then I press decon, say yes to the auto mask and at the default settings I get little white pixels. They vary and even disappear if I alter the settings (e.g. move to "more noise")

By the way, is there a way for DSS to produce autosave TIF (and not FITS) without zip compression? Or, otherwise, perhaps there is a zip library that is easy to integrate so that startools can read those TIF files directly?

Enough to recreate it?

Many thanks for this! Will try to replicate the issue.

ecuador wrote: By the way, is there a way for DSS to produce autosave TIF (and not FITS) without zip compression? Or, otherwise, perhaps there is a zip library that is easy to integrate so that startools can read those TIF files directly?

Definitely! It should be as simple as choosing the 'None' radio button under Compression in the save dialog (and be sure to choose Embed adjustments, rather than apply).

admin wrote:Many thanks for this! Will try to replicate the issue.
Definitely! It should be as simple as choosing the 'None' radio button under Compression in the save dialog (and be sure to choose Embed adjustments, rather than apply).

I was actually asking about the auto-generated autosave.tif. I don't trust the save, it might have been accidentally choosing apply to changes, but it did give me some trouble in the past...

How is it going Ivo, any luck? I've got an image of the Cat's Eye nebula which gives me some very "interesting" results at decon. If I don't set a mask I get the white pixels, but if I set a circular mask around the nebula... well, take a look: That's at default settings, if you want more "fun" try noisier regularization:
Perhaps you'd want the image that exhibits so extreme artefacts, or have you been able to recreate the problems already?

Not good. It's still on my list - the cause is a complex one (Tracking). There are some other issues as well I desperately need time for to look at (notably macOS Sierra incompatibility).

Oh, it's a tracking issue? Well, why don't you create a GUI interface using Visual Basic to see if you can track an IP address??

ecuador wrote:Oh, it's a tracking issue? Well, why don't you create a GUI interface using Visual Basic to see if you can track an IP address??

Bwahahahaha

Legendary piece of TV for all the wrong reasons