Hovercraft sunset

Portsmouth sunset scene from yesterday. With hideous sensor dust :slight_smile:

How can you edit that stuff out? Far as I know RT/DT don’t have that feature. Is there any way of doing it with one click in gimp or something?

Processed with experimental program 3350K, saturation 1.7, mild graduated filter and crop:

My result is a tiny bit more orange than it was in real life. But most software makes it far too yellow.

DSC00319.ARW (23.9 MB)

This file is licensed Creative Commons, By-Attribution, Share-Alike.

5 Likes

The sensor dust is very very easy to remove in DT using the retouch module. darktable 3.6 user manual - retouch

Sorry can’t contribute with the play raw yet at the moment.

here is a quick edit with RT dev 3049. I am quite sure, by spending some more time, the spot removal tool could cover all the visible sensor dust.


Hovercraft_sunset_DSC00319.ARW.jpg.out.pp3 (14,6 KB)

2 Likes

Another great sunset!
Filmulator (default settings) then into GIMP. Sensor dust removed by creating transparent layer, colouring spots with easily seen colour, here green, then select by colour. Changed to the image layer then used Heal-Selection (part of Resynthesizer plug-in).
While dust particles remain constant it should be possible to take a photograph of an out-of-focus white surface to create and save this layer for future use. It is not 100% effective but perhaps a help. The dust spot by the sun still shows.

Possibly a Python script could be written to automate this.

Only slight adjustments made by using luminosity masks.

darktable
DSC00319.ARW.xmp (10.7 KB)

Less yellow-ish with photoflow


DSC00319.pfi (28.1 KB)

thanks for posting
darktable 3.6.1


DSC00319_01.ARW.xmp (61.1 KB)

1 Like


DSC00319.ARW.xmp (17.5 KB)

DSC00319_01.ARW.xmp (17.5 KB)

ART


DSC00319.jpg.out.arp (11.6 KB)

Thanks for posting. I’ve used a somewhat tighter crop. I didn’t want to make it more contrasty or sharper; I liked the soft evening mood. darktable master


DSC00319.ARW.xmp (24.0 KB)

RawTherapee does have a feature to remove sensor defects:

https://rawpedia.rawtherapee.com/Local_Adjustments#Correcting_red-eye_and_removing_sensor_defects

Under RT 5.8-3049 you can apply this sidecar to see sensor stains:
20211123_DSC00319-5.jpg.out.pp3 (52.7 KB)

That sidecar already has the local adjustments done with removing sensor defects according to rawpedia:
Only spot 1 needs to be done and the rest are copies of it, except spot 2 which I have done separately.

The final result without stains and adding a gradient filter:


20211123_DSC00319-4-Result.jpg.out.pp3 (52.8 KB)

Saludos!
Translated with DeepL Translate: The world's most accurate translator (free version)

2 Likes

DSC00319.ARW.xmp (13.4 KB)

Thanks @arturoisilvia @Chen_Hendrawan for explaining rawtherapee and darktable’s dust removal. Wish I’d known when I was using rawtherapee over the last years.

Edit: I couldn’t find rawtherapee’s retouching tool in the official 5.8 release. But I retouched the dust out of my image using darktable, super easy (wow!)

Don’t know how to export in darktable though, so here’s a screenshot of my retouching result:

https://docs.darktable.org/usermanual/3.6/module-reference/utility-modules/shared/export/

In the darkroom, export is in the left-hand panel, at the bottom. In the lighttable view, it’s on the right, also at the bottom.

1 Like

Thanks for posting.


DSC00319.ARW.xmp (31.2 KB)

2 Likes

It is in the development builds:
https://rawpedia.rawtherapee.com/Download

1 Like


GIMP 2.10.24-LAB

Another try at this. Only minor tweaks in GIMP.

My main emphasis was on removing sensor dust and I still missed one (centre top). If I had a “dust” picture (an out-of-focus white background to show the dust particles) and one or two images with the dust, I think I may be able to write a GIMP plug-in to give a one-click clean-up. Unfortunately - fortunate for me - the sensor of my camera is spotless.

As dust spots are out-of -focus they do not have 100% opacity so the original image data still exists. It may be possible to reconstruct the image, but this is way beyond my ability! Perhaps something for the GMIC experts?