Dealing with Dew

[szymon]

So, I've been having real trouble with Dew. Yeah, I know, I should have predicted it and taken steps to prevent it -- I have now, extra dew shield and dew heater on the way -- but I now have images with dew on them that I'd like to work out how best to process. For example, here is a screenshot of my Oiii stack of the Wizard nebula. This is a total of four and a half hours of integration, and I'd really prefer not to throw it away! The Ha and the Sii integration are fine, and I'd like to bring all three together into an SHO image.

So -- what processing strategy is best? Can I load in the Oiii as linear data, somehow tidy it up (the heal module perhaps?), export it out as a tiff (StarTools can't export as FITS, right?) and then import it into the Compose module as Linear Oiii? Would it still be considered 'linear' at that point?

Or is there a way that I can bring it in to the compose module and then process the 'broken' bit while working through it in StarTools myself?

Cheers,

-simon

dew.jpg (491.55 KiB) Viewed 8405 times

[admin]

Assuming that the signal is still usable enough for the purpose of coloring, you can simply use the Compose module like this;
Just load everything as normal (SHO -> RGB), select "L + Synthetic L From RGB, RGB" for "Luminance, Color" and then just simply set Blue Total Exposure to 0 (Not Set).
This will now create a synthetic luminance signal from SII + Ha, leaving O-III out of it. O-III's signal will still be used for the coloring.

Hope this helps!

[szymon]

You know, after parsing that response a few times, I think for the first time I really understand the difference between Luminance and Chrominance in an image. Thank you, that makes sense! I will give it a go (although I'm having different issues with the Sii *sigh* so I might end up doing HOO).

[admin]

You know, after parsing that response a few times, I think for the first time I really understand the difference between Luminance and Chrominance in an image. Thank you, that makes sense! I will give it a go (although I'm having different issues with the Sii *sigh* so I might end up doing HOO).

That's great! Let us know how you go, Personally, I'm always interested in techniques/tricks to still make use of "damaged" data.

[szymon]

Ok, so I tried this this morning, and I am confused. I do this:

confused.jpg (109.91 KiB) Viewed 8354 times

I keep the result (Linear, was not bayered), and then autodev, which gives me a black and white image (without the dew, yay, so clearly only the Ha is being used for the Lum channel, as promised). I then crop out stacking artefacts and go to wipe, and get a nice clean wiped Lum image, but obviously as the Oiii is in the colour then the colour image after wipe shows up the dew marks with wipe halos around them:

wipe_halo.jpg (163.13 KiB) Viewed 8354 times

So surely I'd still have to do something to clear this up to be able to use it even just for colour data, no? I can't even circle it in a mask for wipe until wipe has completed once so that I get the 'Colour' button -- and then circling it doesn't really do anything. What am I missing?

(Also there's a complication -- the autodev for cropping only shows up the Lum channel, which means that I can't see stacking artefacts from the Oiii to crop them. Should autodev have an option to show colour as well as Luminosity?)

-simon

[admin]

Hmmm... Is there anything that prevents you from simply creating a mask for the O-III data beforehand, and importing that into Wipe?

[szymon]

Hmmm... Is there anything that prevents you from simply creating a mask for the O-III data beforehand, and importing that into Wipe?

I'm not sure that I follow...

[szymon]

I've uploaded the stacks here, perhaps you'd be kind enough show me what you mean with a log if you get time

Ha: https://www.dropbox.com/s/7ggnydfy7i9t7 ... .fits?dl=1
Oiii: https://www.dropbox.com/s/mp2gqtauanckb ... .fits?dl=1
Sii: https://www.dropbox.com/s/q77ilv1omi2s1 ... .fits?dl=1

This is 5h25m of Ha, 4h30m of Oiii and 10h25m of Sii. The Sii has no visible nebulosity whatsoever, which is why I'm working on HOO at the moment. It was captured in Light Pollution under a nearly full moon, I'm going to retry the Sii at some point when we get the magic combination of a new moon and no clouds. I was thinking I could use the Sii to add to the Lum layer for the stars perhaps.

Anyway, any ideas you have on processing would be appreciated (maybe I just crop the damned dew spots out and be done with it! ).

-simon

[admin]

I don't think I quite understand what the issue is that you are experiencing...

Are you wanting to create a mask for Wipe that masks out the artefacts/dew spots/circles? And you believe the issue is that you cannot create the mask that Wipe needs to ignore the artefacts?
If so you have many options to create the mask that Wipe needs to ignore the artefacts;

• You can deliberately create an image that shows the circles in all their glory, then create the mask, then use the Restore functionality to restore the dataset to the linear pre-wiped state. This will still keep the mask you just created and it can now be use in Wipe.
• You can load the O-III set, crop and bin it exactly like you would crop and bin the SHO composite, make the artefacts visible with a stretch. Create your mask, save the mask. Now create your SHO composite, bin and crop exactly like you binned and cropped the O-III set, launch Wipe, load the mask you just created.
• You can roughly remember/eyeball the offending circles when creating a mask in Wipe

Remember also that the circles only exist in your colour data; e.g. it requires luminance in the same spot to bring them out.

I gave your dataset a quick process;

--- Compose
Import your SHO data as RGB. On your advice, I used the Ha only as luminance.
Parameter [Luminance, Color] set to [L + Synthetic L From RGB, RGB]
Parameter [Lum Total Exposure] set to [Not set]
Parameter [Blue Total Exposure] set to [Not set]
Parameter [Green Total Exposure] set to [1h00m (60m) (3600s)]
Parameter [Red Total Exposure] set to [Not set]
--- Auto Develop
To see what we got.
--- Bin
Parameter [Scale] set to [(scale/noise reduction 25.00%)/(1600.00%)/(+4.00 bits)]
Image size is 1360 x 912
--- Crop
Parameter [X1] set to [35 pixels]
Parameter [Y1] set to [50 pixels]
Parameter [X2] set to [1331 pixels (-29)]
Parameter [Y2] set to [897 pixels (-15)]
Image size is 1296 x 847
--- Wipe
Did a first Wipe, but noticed I didn't get all stacking artefacts yet in the colour data.
--- Crop
Parameter [X1] set to [70 pixels]
Parameter [Y1] set to [61 pixels]
Parameter [X2] set to [1236 pixels (-60)]
Parameter [Y2] set to [793 pixels (-54)]
Image size is 1166 x 732
--- Wipe
Default parameters. Deliberately not fixing up the dew artefacts.
--- Auto Develop
Default parameters.
--- Color
Makes the dew artefacts visible.
--- Mask
Now I create the mask I need.
--- Restore
Restore data to "Original"
--- Wipe
Vignetting preset.
Parameter [Dark Anomaly Filter] set to [9 pixels].
The masked out artifacts will no longer cause Wipe to locally back off (of course it doesn't remove them, which is also an option).
--- Auto Develop
Parameter [Outside RoI Influence] set to [16 %]
Parameter [RoI X1] set to [388 pixels]
Parameter [RoI Y1] set to [163 pixels]
Parameter [RoI X2] set to [662 pixels (-504)]
Parameter [RoI Y2] set to [586 pixels (-146)]
Parameter [Shadow Linearity] set to [6 %]
--- Color
Tweaked Hubble preset.
Parameter [Bias Slider Mode] set to [Sliders Reduce Color Bias]
Parameter [Style] set to [Artistic, Not Detail Aware]
Parameter [LRGB Method Emulation] set to [RGB Ratio, CIELab Luminance Retention]
Parameter [Dark Saturation] set to [5.30]
Parameter [Bright Saturation] set to [2.00]
Parameter [Saturation Amount] set to [200 %]
Parameter [Blue Bias Reduce] set to [3.59]
Parameter [Green Bias Reduce] set to [14.70]
Parameter [Red Bias Reduce] set to [1.00]
Parameter [Cap Green] set to [100 %]
--- Wavelet De-Noise
Parameter [Grain Size] set to [3.4 pixels]
Default parameters.

NewComposite.jpg (256.98 KiB) Viewed 8331 times

Hope this helps!

[szymon]

Ohhh! I see, you mean using time travel. Ok that makes sense

This bit was interesting:

The masked out artifacts will no longer cause Wipe to locally back off (of course it doesn't remove them, which is also an option).

How would you get wipe to remove them?