indexes to solve narrow band of images

[Aaron]

Dear astrometry.net community,

I have very narrow band of images for which I am trying to solve the pattern (roughly 0.1deg x 1deg). 

Looking at the documentation for the available database indexes, I feel like I might need to build custom indexes with skymark diameters even smaller than xx00 series (2-2.8arcmin).

In your experience, would you agree with the above? Would you have an order of magnitude of how small the skymark diameter needs to be with regard to the shorter field-of-view side?

Thank you so much in advance!

Regards,

Aaron

[Icy Kofan]

Dear community,

I have not been able to solve a pattern yet on these narrow bands of images and I just wanted to confirm with you whether this would be caused by the very rectangular shape. In the python library interface for instance, we can only provide a single field-of-view value (radius around the center). That would mean that the solver expects stars pattern covering that full area rather than a subset of it right? in the case of a very rectangular shape, if the radius is then the longer side, and that there are only star coordinates in a narrow band, would that be an issue?

Thank you in advance for any input you might provide! :)

Regards!

[Dustin Lang]

Hi,

Would you mind sharing one of your images?  You could, for example, upload it to nova.astrometry.net and send the link to the submission.

Thanks,

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/80b0a1bf-920e-42ea-a1ba-a223f406ad84n%40googlegroups.com.

[Icy Kofan]

Hi Dustin, 

Thanks for your answer, the png should have been uploaded with name aaa_0.png.

Astrometry.net submission 7087721

Aaron!

[Dustin Lang]

That looks like a screen-shot of a plot of the data... can you upload the original data file?  (FITS?)

Trying to solve an image like that -- with tick marks and a lot of white space -- is very unlikely to work :)

cheers,

dustin

To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/41f0f433-51d9-4b04-8b6b-de2f57a4624bn%40googlegroups.com.

[Icy Kofan]

oops, my bad, I exported to the wrong format. That should be better Astrometry.net submission 7087779

[Dustin Lang]

Ah yes, much better :)

It looks like these images have a fair number of cosmic rays, and pretty few actual stars (and a weird PSF!).

I think that's going to be pretty tough to get results out of.  Our built-in star detector is not smart about cosmic rays, so it will find many of those and treat them as "stars".  So if you can remove them (eg, with a median filter), that will help.  Normally I would say that you need 10-15 stars at a minimum to get a solution, but with your very rectangular fields, it's probably going to be even more than that.

cheers,

dustin

To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/813ccc28-45ca-48cb-a2a0-7e1985cb3992n%40googlegroups.com.

[Icy Kofan]

I have tried passing in a TXT files as well with stars in it instead of the image (that's how I use it internally). But unfortunately without success as well.

Another image with a few more stars yielded results though.

I thought that this one was solvable but so far, it has failed with different index scheme (triangle or quads).

What you are seeing isn't really a PSF though. It's just related to the movement of the sensor.

Thanks again!