image does not solve locally but solves quickly at nova.astrometry.net

[mlco...@gmail.com]

I have an image that solved quickly at nova.astrometry.net, job 8224308, but it does not solve locally.

I wonder what the difference could be.

The field is fairly small, 6.93 x 4.65 arcmin.

And my index files are from here: https://portal.nersc.gov/project/cosmo/temp/dstn/index-5200/LITE/

And I'm using the Debian package astrometry.net for solve-field

Thanks.

Mark

[Dustin Lang]

Hi,

It looks like the web version solved it with index file:

Field 1: solved with index index-5200-22.fits.

The web version does the equivalent of "--downsample 2", maybe that will help?

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/c7f24f0f-d0f9-4608-b48d-13277a6a5e71n%40googlegroups.com.

[Dustin Lang]

Hi Mark,

"--downsample 2" changes how stars are detected -- it bins the image 2x2 before running the star detection algorithm.  If your solve-field is making plots, you can look at the *-objs.png file with and without downsampling and see which stars are detected.  Downsampling tends to work better if stars are several pixels wide, and especially if they're saturated -- which is why it's the default for the web, and maybe should be the default for solve-field!

cheers,

dustin

On Sat, Apr 8, 2023 at 1:21 AM Mark Copper <mlco...@gmail.com> wrote:

Thank you, Dustin. That worked.

But why, why, why did it make a difference?

Mark

[Mark Copper]

Thank you, Dustin. That worked.

But why, why, why did it make a difference?

Mark

On Fri, Apr 7, 2023 at 8:00 PM Dustin Lang <dstn...@gmail.com> wrote: