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

My Issues with SGP Plate Solving were my bad

  • Please log in to reply
1 reply to this topic

#1 Peregrinatum

Peregrinatum

    Viking 1

  • *****
  • topic starter
  • Posts: 804
  • Joined: 27 Dec 2018
  • Loc: South Central Valley, Ca

Posted 14 February 2020 - 12:40 PM

Couldn’t figure out why SGP wasn’t plate solving with the C925HD at F10… when SGP plate solving works it’s like caffeinated coffee, when it doesn’t it shuts you down for the night and you go on CN and complain about it…  I was tired of complaining and did some digging around and I had a few issues going on that I hadn’t realized yet, my bad.

 

First issue is that Plate Solve 2 will only work if the mount slews within a few degrees of the target, and at F10 I had a larger margin of error… 

 

In addition it seems I didn’t configure ASNVR (local Astrometry.net plate solver for Windows, found here:  https://adgsoftware.com/ansvr/) properly when I installed it, and it was missing needed index files for blind solving at F10…

 

Here are a few things I learned that really helped, hope this makes sense and helps you:

 

[1] “Center Now” to start the imaging session - SGP won’t be successful with Plate Solve 2 (PS2) unless your initial slew is within a few degrees of the target.  If your mount’s pointing model is off, or not very accurate at a higher FR like F10 as mine was, then PS2 just won’t work, not it’s fault that’s just the way it is designed…  When the FOV starts to decrease with a larger FR this becomes critical.  So if PS2 won’t solve then you try to blind solve and sync, and that will usually jump start PS2 and get it working.

 

[2] “Slew” to target and Blind Solve and Sync with ANSVR, then “Center” with PS2 - This works IF you have the correct indexes installed for ANSVR to use.  This was key for me to blind solve at F10, I was missing the needed index files that are required with a smaller FOV.  I just assumed the “standard” install of ANSVR would work fine, and I was careless when I installed the index files…  this is the window I am referring to during the ANSVR installation:

  

j9dpLmol.jpg

 

 

[3]  You need to figure out which ANSVR index files are needed for the FOV of your imaging train (consider plate scale and camera sensor dimensions).  Lots of online helps to guide you in figuring out your imaging train FOV, and it needs to be in arc-minutes not degrees (multiply degrees by 60).  For example, my FOV using the C925HD F10 with the ASI1600MM is 19.5 X 25.9 arc-minutes.  If you dig into ANSVR user’s manual (found here: http://astrometry.net/doc/index.html) they suggest you use the index file that would cover half your smallest FOV dimension (this is my interpretation), so for my setup I need the index that covers 0.5*19.5 ~ 9.8 arc-minutes… below is a list of the index files, so I would need index file 4204.  When my plate solves were failing at F10 I was missing this file (I only had files 4207 to 4219), makes sense now why ANSVR wasn’t blind solving…  Here is the link to the index files:

 

http://data.astrometry.net/4200/

 

vb6LFTSl.jpg

 

[4]  ANSVR used to take forever to plate solve, now it takes 10sec… here is what I learned to make it run smoking fast.  Only load the index files you actually need for the particular FOV you are currently imaging at.  Go to the location where the index files are stored on your computer, mine are located here:

 

C:\Users\YOUR_USERNAME\AppData\Local\cygwin_ansvr\usr\share\astrometry\data

 

Create another directory and move all the files to that directory, then copy only the index files you need to the location listed above…  this way ANSVR will only search the needed indexes and not all of them…  so for my example I would only copy the 4204 files to that directory.  The download for the ANSVR index files are located at this url:  http://data.astrometry.net/4200/

 

There are a few settings that you should use so that ANSVR runs at it’s fastest, especially the error estimate = 5 and objects = 100.  You can access the ANSVR configuration page in SGP using the plate solve tab in the control panel.

 

JwL6Dtql.jpg

 

[5]  You can verify what index you will need by uploading an image to the http://nova.astrometry.net/ to have it plate solved.  Or, if you find an image on astrobin you could upload it as well to give you an idea which index is needed… Here is an image I uploaded, a single frame 30s NB Blue filter image of the Eskimo Nebula, and when the plate solve was successful it showed which index it used… it is hard to see in this pic but the circled field says it used index 204, this corresponds to ANSVR index 4204, the same one I intended to use.

 

LMVETALl.jpg

 

 

Here is my new workflow now for plate solving in SGP:

-determine which ANSVR index files are needed for the FOV I will image at
-load that index only into the ANSVR index file directory
-“Slew” to target in SGP, when it stops take a pic using the framing tool
-“Blind Solve and Sync” the image using the plate solve tab in the control panel
-“Center” to the target with PS2 and start imaging


Edited by Peregrinatum, 14 February 2020 - 01:30 PM.

  • dswtan, jdupton, RossW and 1 other like this

#2 ZL4PLM

ZL4PLM

    Vendor - ZWO Product Dev Team Member

  • *****
  • Vendors
  • Posts: 1,390
  • Joined: 23 Apr 2017
  • Loc: Darfield, New Zealand

Posted 14 February 2020 - 03:08 PM

I've ditched ps2 now ...using astap much faster less failures with sgp




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