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

Could not get on well with Star registration in PI....bought something else

  • Please log in to reply
10 replies to this topic

#1 pyrasanth

pyrasanth

    Skylab

  • *****
  • topic starter
  • Posts: 4,021
  • Joined: 08 Jan 2016

Posted 20 April 2021 - 11:19 AM

I don't know what the difficulty is but PI seems very picky about what settings are selected in order to register data from widely differing image scales- I tried previews, thin splines and nearly every combination of settings I could think of and mixed images would not register. Image from the same camera okay but mixed images not having it.

 

However I found a great software suite that I've just purchased which is solely dedicated to star registration and processing. It is a bit quirky and the interface could be a bit more modern however it works really well. Just set any image of any image scale as the master (even OSC bayered) and it works against a full folder of 16 bit images and churns through them in sequence. It even calculates and corrects the image distortion between different optical systems and compensates forthwith- so I've read in the help files. 

 

I also had an earlier copy of CCDStack however this software really makes that look like a freebie!

 

I'm using this software to build my NGC3718 image as there are 2600 subs imaged through 3 camera and 2 telescopes with widely differing image scales. The software costs $175.00 however it is well worth the money. You can get a trial copy to play around with however its not much use as you can't save any output so you need confidence that it works and bite the bullet as they say. The only small limitation is that the software works at 16 bit- it is supposed to correct 32 bit images to 16 bit truncation but I could not get that to work so I've resampled my entire data stack to 16 bit.

 

I've only tested a small data subset of 20 of the registered images however its looking very good as the sample came from 3 different cameras and I see no differential distortion verses distortion seen at the edges of the masters when I tried to integrate the 3 seperate camera stacks through PI.

 

The link is here https://aurigaimagin...al/ordering.php


Edited by pyrasanth, 20 April 2021 - 11:22 AM.


#2 yzhzhang

yzhzhang

    Apollo

  • *****
  • Posts: 1,039
  • Joined: 13 Dec 2012
  • Loc: SoCal

Posted 20 April 2021 - 12:52 PM

Maybe also worth checking AstroPixelProcesser (APP). We use it mostly for pre-processing including calibration and registration. Works very well especially for merging data with different scales and/or compositing Mosaics.


  • lambermo and Lead_Weight like this

#3 microstar

microstar

    Surveyor 1

  • -----
  • Posts: 1,529
  • Joined: 05 Jan 2008
  • Loc: Nova Scotia, Canada

Posted 20 April 2021 - 12:59 PM

RegiStar has been around for ages. Does one thing but does it very well.

#4 Rasfahan

Rasfahan

    Viking 1

  • -----
  • Posts: 563
  • Joined: 12 May 2020
  • Loc: Hessen, Germany

Posted 21 April 2021 - 02:57 AM

Thanks for the heads up. I have problems registering a multi-night-project of NGC4565, taken with 2 scopes and 3 cameras. About 20% of the subs will not register at all, and I can see no difference to ones that register. Tried different parameters and registration targets, that just changes which frames do not work. I have a copy of APP and will try that first. If that fails, too, Registar may be the rescue.
  • pyrasanth likes this

#5 Dan_I

Dan_I

    Vostok 1

  • -----
  • Posts: 177
  • Joined: 22 Sep 2010
  • Loc: France

Posted 21 April 2021 - 08:19 AM

Did you try the freeware Siril, which is easy to use and produce consistently good results ?



#6 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,987
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 21 April 2021 - 08:47 AM

I second, Astro Pixel Processor. I've used it to combine multiple sessions from multiple telescopes and cameras. It will scale the images to match. Here's one I did mixing Edge 11 data with a 102mm refractor, the ASI2600MC-Pro with ASI1600MM-Pro.

 

get.jpg?insecure

 

Then here's a massive integration I did with 27 hours containing 543 60s color subs and 2643 30s mono subs and another hour or so of HA data. It was a processing nightmare that took around 16 hours to stack and calibrate. It was a mix of mono and color data from the ASI071 and ASI1600, but shot on the same telescope.

 

get.jpg?insecure

 

APP handled both of these without issue though I did have it set to use max memory, and processed on an 8 core xenon.


Edited by Lead_Weight, 21 April 2021 - 08:56 AM.


#7 Rasfahan

Rasfahan

    Viking 1

  • -----
  • Posts: 563
  • Joined: 12 May 2020
  • Loc: Hessen, Germany

Posted 21 April 2021 - 09:37 AM

Nice images. The problem is not RAM or processing power (92GB RAM VM on a Threadripper 3960X). I used APP some, but I‘ve had some crashes in the past which is annoying since you can not save your projects (light/dark/flat assignments). WBPP can save them even when things go south. When you have sorted through several thousands of flats, flat-darks, lights and darks losing it all because of some software glitch is annoying. Not getting all subs registered is, too, of course.



#8 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,987
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 21 April 2021 - 09:39 AM

Nice images. The problem is not RAM or processing power (92GB RAM VM on a Threadripper 3960X). I used APP some, but I‘ve had some crashes in the past which is annoying since you can not save your projects (light/dark/flat assignments). WBPP can save them even when things go south. When you have sorted through several thousands of flats, flat-darks, lights and darks losing it all because of some software glitch is annoying. Not getting all subs registered is, too, of course.

Yeah, this was definitely a problem in the past. And when I made the M101 image, this happened a few times because I had not increased the RAM allocation. Unfortunately it doesn't do dynamic RAM allocation, so setting it to max alleviates this issue. And I've never had a crash since. I believe the developer has also reworked a few things to alleviate crashes, and the next version of APP is supposed to save sessions.



#9 Rasfahan

Rasfahan

    Viking 1

  • -----
  • Posts: 563
  • Joined: 12 May 2020
  • Loc: Hessen, Germany

Posted 21 April 2021 - 09:42 AM

Great, I can wait for next version, get some more data...



#10 gatsbyiv

gatsbyiv

    Apollo

  • *****
  • Posts: 1,033
  • Joined: 29 Mar 2009
  • Loc: Doylestown, PA

Posted 21 April 2021 - 10:33 AM

APP is pretty impressive in its ability register pretty anything you throw at it.  And it keeps everything in 32 or 64 bit; that's a little concerning that Registar drops data to 16 bits.  

 

PI's WBPP is a big step forward, but I agree that its star registration and image solving abilities seem to have fallen behind the competition. 



#11 SeattleStarGazer

SeattleStarGazer

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 18 Oct 2007

Posted 24 April 2021 - 02:58 PM

I love APP, it's been very reliable lately with me setting RAM allocation to near Max (48GB out of 64), but need a better computer to speed things up. I have a Dell Precision Workstation from 4-5 years ago with Xeon processor and Quadra video with dual PCIe SSDs and it's been great until I started going with CMOS and theQHY 268C - hundreds of 26-MP files and taking 12 hours to completely preprocess. All automatic mind you but I hate the wait to see my results! The APP guys are recommending more cores, what about the AMD Ryzen line? Any specific recommendations?

 

Wayne




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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics