Persistence of Search Radius

[Wade Hilmo]

I am using ASTAP 0.9.459 64 bit as my solver with SGP. I further have Astro-Physics Point Mapper (APPM) doing a sky model run through SGP.

I am having two issues.

The first issue is that solves occasionally fail. That is not the issue I'm running down right now. The failures are infrequent and generally not a problem,

The second issue is that when a solve fails, it completely breaks the modeling run. The issue is that APPM times out the solve attempt while ASTAP is still searching. The long search is because early on, when I was first testing ASTAP, I experimented with blind solving and had set the search radius to 90 degrees in the app. The problem now is that, no matter what I try, I cannot change the search radius in a persistent way.

I have tried using "Save Settings". I have tried doing "Save Settings As" and then reloading the settings. I have tried changing the settings running ASTAP from the start menu as a graphical app, and I have tried changing the settings by using the properties button in SGP to create the ASTAP process. In all cases, I am trying to set the search radius to 10 degrees.

Sometimes, if I exit and restart ASTAP, I will see that the search radius is 10 degrees on the new instance of ASTAP. Sometimes, if I exit and restart, I will see that it's reverted to 90 degrees. 100% of time time, if an APPM solve fails, it is using a 90 degree search radius.

Where is the setting actually getting stored? How can I eliminate the 90 degree search radius from every possible source (I suppose that SGP could be including a -r parameter when it creates the ASTAP process, but this would be surprising to me, since it doesn't expose that as an option, and it's a terrible default).

Thanks for any suggestions.

[han.k]

Hello Wade,

The search range is defined by SGP and passed on to ASTAP by the command line. This will overwrite the ASTAP setting. SGP still execute ASTAP as is it Platesolve2 so it is expressed in regions rather then degrees. You have to set the number of regions in SGP to a lower value.

Han

[Wade Hilmo]

Thanks for the response.

Is there any way to map the Search Regions in SGP to the actual Search Radius value that ASTAP will use?

[han.k]

If SGP would execute ASTAP with the native command line then it would possible to define a search radius in degrees. But unfortunately they still use the platesolve2 command line for ASTAP. This is roughly converted to a radius in degrees.

Assume you specify 1000 regions, the is equal to about 32 x32 =1000, so a radius (from the center of 16 areas. If your image height is 1 degrees, the search radius will be 16x1=16 degrees. 4000 regions will result in a radius 32 fields ro 32x1 =32 degrees and 16000 regions in a radius of 64 fields.

In a formula :

radius:= image_height x sqrt(regions)/2

[Wade Hilmo]

I have a small field, at 21' x 17'. Using this, it looks like there is a huge difference between 2500 (the largest number SGP offers) and "Max Regions".

If I select 2500, I believe that the search area will be approximately 14 degrees by 9 degrees. Based on what I've seen while the modeling run is happening, ASTAP will fail a 14x9 degree search long before APPM times out. The next step up from there is "Max Regions", which uses a 90 degree radius and can take a very long time to fail.

Thanks for the help!
-Wade

[han.k]

With 2500 regions your search range will be about a square 7 degrees.

For 21x17' you better have the G18 star catalog installed. The G17 has not enough stars. Remove the G17 to be sure it selects the G18.

Clear skies, Han

[han.k]

Note ASTAP version 0.9.449 or later will improve the solving speed of your small field.

[Wade Hilmo]

Thanks for the tips. I've switched from the G17 to G18 catalog, and I have the latest version of ASTAP.

I also kicked off another APPM run tonight with SGP set to 2500 search regions. I was able to confirm that this addresses my issue. When a solve failure happens now, ASTAP stops searching before APPM times out, so everything recovers fine.