Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

StarNet v2

Astrophotography Imaging Software
  • Please log in to reply
328 replies to this topic

#326 kopi

kopi

    Lift Off

  • -----
  • Posts: 2
  • Joined: 18 Jul 2022

Posted 25 September 2022 - 08:05 AM

Those instructions seem Pixinsight-specific, as they're placing libraries in the Pixinsight path, rather than the StarNet one (which is fine if you're using Pixinsight, less so if you're not).

 

These instructions appear to be more applicable:

https://www.williaml...om/starnet-cuda

Sorry for the late response. I got a little distracted in the past weeks.
Today I checked the instructions you've linked and they work beautifully.

 

Thank you, kopi


  • JeffreyHorne and AaronH like this

#327 Glass17

Glass17

    Vostok 1

  • -----
  • Posts: 164
  • Joined: 02 Feb 2022

Posted 02 October 2022 - 04:48 PM

I've been enjoying StarNet2 in PixInsight, but noticed that if I apply it after I've run background extraction I get a check pattern in the white area of stars, particularly noticeable in larger stars. I avoid this by applying StarNet2 immediately after I crop and before running any other processes.

But I still have a problem. Sometimes I like to stack in Astro Pixel Processor and use its tools to crop, fix vignetting and light pollution and do background calibration and star color calibration. When I bring the resultant fits file into PixInsight and apply StarNet2, I get the check pattern in stars again.

 

I'm able to de-emphasize the pattern with PI l Convolution and Curves, but don't prefer this approach.

I'm hoping for a fix to this because the image from APP edits nicely in PI, but I don't want to do without StarNet or try a competitor.

Please advise.

I got the same and even reported it in my post from August 12.

Now I think I was able to  found where it is coming from - for me it is coming from drizzling.

If some quality advantages  even coming from drizzling (not much I can see in real life) 2x2 drizzling creates this grid pattern and slow down any further processing more then four time (I think five-six times at least). Going to give up on this and focus my time more on improvement other processing techniques.


 



#328 phtnnz

phtnnz

    Sputnik

  • -----
  • Posts: 28
  • Joined: 11 Oct 2018

Posted 02 October 2022 - 05:24 PM

OSC data and Bayer drizzle? This yields similar artifacts around bright stars even without using StarNet2, at least in my case.

I switched to the new WBPP and separate processing of the R, G, B channels to avoid these problems.



#329 Glass17

Glass17

    Vostok 1

  • -----
  • Posts: 164
  • Joined: 02 Feb 2022

Posted 02 October 2022 - 08:40 PM

OSC data and Bayer drizzle? This yields similar artifacts around bright stars even without using StarNet2, at least in my case.

I switched to the new WBPP and separate processing of the R, G, B channels to avoid these problems.

In my case mono and pysiril (Siril)
cmd.register(StackName, prefix="r_", drizzle=True)

It leads to check grid late with Starnet2 on COREs of the bright stars usually on bright phone of nebulas.
Take a look on the photo in my post here from August 12 2022

For pre-processing I prefer Siril and for processing I am going to PI


Edited by Glass17, 02 October 2022 - 08:41 PM.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: Astrophotography, Imaging, Software



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics