Dusty Taurus reprocess with Starnet

[Allinthehead]

Just to test Starnet star removal on some old data. I ran the script in PI and brought it into PS. I used a less stretched version and layered it into the starless version in blend mode lighten. I then layered the version i had previous to the stars being removed and layered that back in again and played with the opacity until i had something decent.

This is the result after a curves layer.

 

 

And this is the original.

To my eye at least the stars are significantly reduced and the more dust has been teased out.

 

[cuivenion]

Both the images are great, diffraction spikes are greatly reduced in the starnet image. Thanks for posting about starnet, I hadn't heard of that. I tried using Stratton but it did weird things to the stars.

[Allinthehead]

10 minutes ago, cuivenion said:

Both the images are great, diffraction spikes are greatly reduced in the starnet image. Thanks for posting about starnet, I hadn't heard of that. I tried using Stratton but it did weird things to the stars.

Starnet works very well. 

[Allinthehead]

Another example here.

New version

 

Original

 

Star removed

[Laurin Dave]

Very nicely done Richard....  StarNet's a great tool

Dave

[Brian Maurer]

You do a great job dealing with the stars, that it’s hard to choose.  I think in the first images, they both look good. In the 2nd the original looks better. Somehow the stars in the “reduced” version don’t seem to have the color or pop. Again, I think it’s because you did such a good job on the originals. 

[gorann]

Very interesting Richard. I heard of Starnet but have not tried it yet. As Brian says, you did such a good job on the "originals" that the effect is not so obvious, but it is there. Will you use it routinely from here on and maybe apply it earlier in the process (e.g. remove the stars before any major stretching)?

[Allinthehead]

52 minutes ago, Brian Maurer said:

You do a great job dealing with the stars, that it’s hard to choose.  I think in the first images, they both look good. In the 2nd the original looks better. Somehow the stars in the “reduced” version don’t seem to have the color or pop. Again, I think it’s because you did such a good job on the originals. 

Thanks Brian. I suppose the object of the exercise was to demonstrate how effective Starnet is. I think i spent about 10 minutes processing these so with my normal routines i reckon i could improve on these still.

 

21 minutes ago, gorann said:

Very interesting Richard. I heard of Starnet but have not tried it yet. As Brian says, you did such a good job on the "originals" that the effect is not so obvious, but it is there. Will you use it routinely from here on and maybe apply it earlier in the process (e.g. remove the stars before any major stretching)?

Yes the effect is subtle until you blink between the two versions and it becomes very obvious. According to the author Starnet is best used on already stretched images, so what i may try next is to do a crazy version and remove the stars and then a version just concentrating on star colour and size, blending the two to taste.