too many sources?

[Noy Shitrit]

Hello,

I have some fields that i have trouble solving 

The field size is 15.9x15.9 arcminutes 

When solving - using the 4205-* index files and 4206-*

it apears to me as if it detects too many sources and then - either gets to the CPU limit or either get stuck in the middle (if i increase the CPU limit) without writing any solution or summery.

I tried as well, to change the sigma and the downsampling - tough i am not sure if im loosing important information there 

When solving it in the nova.astrometry.net it sucseeded to do it only once "using SExtractor", the other times it reaches the CPU limit and failing to solve (even if i try use SExtractor again)

I would be grateful for your help.

attaching a link for one of the fits files i am working with - https://drive.google.com/file/d/1ZlloueQgkaq7OtoBkliY4i2zvhY71P7b/view?usp=sharing

Noi

[Dustin Lang]

Hi,

As you have probably noticed, this image looks like it might be out of focus -- the stars are 'donuts'.  That's usually tough for source finders to work with!  I got a good result using nova.astrometry.net with Downsample = 8 -- that seems to be good enough to turn the donuts back into something that can be centroided.

http://nova.astrometry.net/user_images/3558941#annotated

cheers,

--dustin

--
You received this message because you are subscribed to the Google Groups "astrometry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to astrometry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/2fc060b8-0513-4b01-bf5c-47da1196918d%40googlegroups.com.

[Noy Shitrit]

Hey again dear Dustin, 

thank you for your help.

thats right, my images are defocused (on purpse) 

you right, the last image get solved with downsampling 8

the problem is that i keep getting stuck on other images, for example the one in this link (tried also to "play" a bit with the parameters) 

https://drive.google.com/open?id=1ALAabbOxxIIP1XOaJz-yFvSi5yg9-N8l

this is the command line i ran for this image:

/usr/local/astrometry/bin/solve-field  coj1m011-kb05-20140430-0499-e90.fits --overwrite --downsample 8 --ra 221.607, --dec -37.229, --radius 0.18

when i was changing the downsampling in the range of 2 - 12..

just a note, i have over 6000 files thats needs to be solved changing manually the parameters for every image wont be the best way

On Wednesday, April 15, 2020 at 3:28:47 PM UTC+3, Dustin Lang wrote:

Hi,

As you have probably noticed, this image looks like it might be out of focus -- the stars are 'donuts'.  That's usually tough for source finders to work with!  I got a good result using nova.astrometry.net with Downsample = 8 -- that seems to be good enough to turn the donuts back into something that can be centroided.

http://nova.astrometry.net/user_images/3558941#annotated

cheers,

--dustin

On Wed, Apr 15, 2020 at 8:04 AM Noy Shitrit <noysh...@mail.tau.ac.il> wrote:

Hello,

I have some fields that i have trouble solving 

The field size is 15.9x15.9 arcminutes 

When solving - using the 4205-* index files and 4206-*

it apears to me as if it detects too many sources and then - either gets to the CPU limit or either get stuck in the middle (if i increase the CPU limit) without writing any solution or summery.

I tried as well, to change the sigma and the downsampling - tough i am not sure if im loosing important information there 

When solving it in the nova.astrometry.net it sucseeded to do it only once "using SExtractor", the other times it reaches the CPU limit and failing to solve (even if i try use SExtractor again)

I would be grateful for your help.

attaching a link for one of the fits files i am working with - https://drive.google.com/file/d/1ZlloueQgkaq7OtoBkliY4i2zvhY71P7b/view?usp=sharing

Noi

--
You received this message because you are subscribed to the Google Groups "astrometry" group.

To unsubscribe from this group and stop receiving emails from it, send an email to astro...@googlegroups.com.

[Dustin Lang]

I don't have any good advice for getting consistent centroids for donut images.

As you can see, our built-in source detector gets the peak incorrect -- the bright stars at the top here are mis-centered --

http://nova.astrometry.net/extraction_image_display/4123647

This is because the source detector is using a Gaussian matched filter, and the donut is a really poor match for that!

Maybe others in this group have experience handling images like this...?

--dstn

To unsubscribe from this group and stop receiving emails from it, send an email to astrometry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/67da1fbb-04c3-451c-af57-80bbe2f1e625%40googlegroups.com.

[Noy Shitrit]

Hey,

what if i run my files first in SExtractor (woth a tophat filter for example), gets an XYlist and then put it to astrometry.net to fix the deviation from the real RA DEC?

To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/67da1fbb-04c3-451c-af57-80bbe2f1e625%40googlegroups.com.

[Dustin Lang]

Yes, that would be an option.  If you have a source extractor config file, you can run it via solve-field:

solve-field --sextractor-config my.conf   [..usual args..]

To unsubscribe from this group and stop receiving emails from it, send an email to astrometry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/de88378d-9406-4aaf-bbe2-f1202aeff858%40googlegroups.com.

[Noy Shitrit]

Hello dear Dusting again,

regarding the downsampling of the images,

I want to make sure,  are the new image files created after the engine solved the image are as seen also after the downsampling process? or it is using it only to solve the image and then "going back" to the original image just with the right astrometry solution?

Thank you,

Noi

[Dustin Lang]

Hi,

The second option you said -- the downsampled star positions are used for solving, and then the solution is scaled back up to apply to the original image.

cheers,

--dustin

To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/33d4afac-8b34-418a-a4b0-ad9561ed96f8n%40googlegroups.com.