StarTools

Hi all,

I'm not very knowledgeable in ST so wondered if you can tell me how to remove the black grittiness on the background of this image? (JPEG - Reduced) I used Dev only, as Autodev made the image too noisy. I cropped out stacking imperfections which made Dev a lot better, but the noise is difficult. The image I've attached has just had Dev. Nothing else, as I wanted you to see the grittiness.

I was using an Atik 314L+ and find adding darks to images just makes it worse, for some reason. So there are some dead pixel artifacts.

17 x 8 Min subs in Ha. Cooled.

Can you help?

Cheers, Alex

There are several ways you could approach this (non-expert disclaimer):

After using wipe and upping the Dark anomaly slider to, say 5 or 6, go to Autodev, select a region of interest (ROI) and then adjust that to get a better balance between the nebula and the background, then adjust the ignore fine detail slider, try around 2 - 3% for starters. When you have found the best balance there, try adjusting the Outside ROI slider down to 4 or 5%.

You can also use the Life module, Isolate preset with maximum Airy disk radius and Mask fuzz values, although probably at less than full strength.

After HDR, Sharp and Decon, use the Life module again in Isolate mode but this time set the Airy disk radius to 3 or 4 pixels and adjust the strength to taste.

Finally, after the colour module near the end, at the turn off tracking point adjust grain size to get rid of any residual noise grain.

Is there any chance you could post a link to the original stacked image on Dropbox or similar? We could have a play then and see what works and what doesn't with this particular image.

The original stack would indeed be most helpful!

The hot pixel trails indicate that you did not dither between frames. Dithering, after flats, is the most effective (and usually cheapest!) way you can improve your datasets immensely and avoid things like pattern noise, which is what we are presumably talking about here ("black grittiness"); noise grain should be one pixel in size and not bleed into neighboring pixels. Noise, when zoomed in, should not look like streaks, blobs or "worms".

Dithering, combined with an outlier rejection stacking algorithm, will also eliminate hot and dead pixels.

Many thanks!