Solve-field works on Website but Fails Locally

102 views
Skip to first unread message

Dong Yun Jung

unread,
Feb 1, 2025, 12:58:37 PM2/1/25
to astrometry

Hi everyone,

I'm having trouble getting solve-field to work. It worked on the website, but I installed it via conda-forge on Anaconda and also installed the index file referenced in the website logs(I downloaded all of 5200~5206, homepage says solved with 5203-10 ). My logs indicate that the index file is recognized and sources are detected. However, when I run the following command in the terminal:

solve-field  ~/ccd_1.fits --ext 1 --dir . --continue --scale-low 1 --scale-high 14 --scale-units amw

it doesn't work. By the way, the FOV is approximately 6 arcmin by 14 arcmin.

Thanks in advance!

Dustin Lang

unread,
Feb 1, 2025, 4:03:03 PM2/1/25
to Dong Yun Jung, astrometry
Hi,

You might try adding "--downsample 2", which the web site does by default.

The "amw" option for scale-units means "arcminutes wide", so if your images are 6 wide by 14 high (tall images), that should be fine... if they are 14 wide, then you'll want to increase --scale-high.

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 visit https://groups.google.com/d/msgid/astrometry/653aeef3-9862-4895-a51c-6e59297d906dn%40googlegroups.com.

Dong Yun Jung

unread,
Feb 2, 2025, 3:37:59 AM2/2/25
to astrometry

Thank you for your help.

My image is indeed a "tall image" as you suggested. However, even after adding --downsample 2 to the command, it still doesn't work. Most of the execution log is filled with lines like:

Field 1 did not solve (index index-5203-12.fits, field objects 1-10).
Field 1 did not solve (index index-5203-11.fits, field objects 1-10).
Field 1 did not solve (index index-5203-10.fits, field objects 1-10).
Field 1 did not solve (index index-5203-09.fits, field objects 1-10).

and the final line states:

Did not solve (or no WCS file was written).

For additional information, here is the link to the image that was solved on the website:
https://nova.astrometry.net/user_images/11910373#annotated

Thank you again for your assistance.


2025년 2월 2일 일요일 오전 6시 3분 3초 UTC+9에 dstn...@gmail.com님이 작성:

Dustin Lang

unread,
Feb 2, 2025, 9:02:30 AM2/2/25
to Dong Yun Jung, astrometry
Hi,

Okay, that's unexpected!  Can you please add "-v" to your solve-field command and send the full long output that results?

If your local version has plotting, try adding the arg "--plot-scale 0.25" and then check that the result looks more-or-less like the web version --

In the web solve logs, it finds the answer after object 10 and it gets a very strong result -- this image is "easy" :)

thanks,
dustin

Dong Yun Jung

unread,
Feb 2, 2025, 1:42:02 PM2/2/25
to astrometry
Hello,

Thank you for your quick reply. I uploaded the log. You  said the image is "easy", so i hope it will be solved easily.

Thank you.

2025년 2월 2일 일요일 오후 11시 2분 30초 UTC+9에 dstn...@gmail.com님이 작성:
solve_log.txt

Dustin Lang

unread,
Feb 2, 2025, 3:03:32 PM2/2/25
to Dong Yun Jung, astrometry
Hi,

I'm not sure why this is happening, but I see it printing:

Image size: 8 x 4644

I'm guessing your images are compressed with .fz / fpack.  The astrometry.net tools are not smart about handling that.  It does looks like the source detector figured out the correct size of your image, but later it got confused.

Can you try either:
- use funpack to uncompress the image to a temporary file and try that; or
- try adding the arguments "--width 2112 --height 4644"

cheers,
dustin




Dong Yun Jung

unread,
Feb 3, 2025, 12:55:17 PM2/3/25
to astrometry
Hello,

After decompressing by using funpack, it works!! Thanks a lot!!

2025년 2월 3일 월요일 오전 5시 3분 32초 UTC+9에 dstn...@gmail.com님이 작성:
Reply all
Reply to author
Forward
0 new messages