Which stars are used? objs vs. xyls

[Max van de Kamp]

I am using the astrometry c code on my own computer. 

When I run solve-field, one of the first things it does is make a copy of the original image '-objs.png', with circles around many of the stars. I was assuming that those are the stars which it will try to track down. But in the ultimate 'xyls' file, the x-y-positions are not the same as in that picture. Some are the same, but many are added, and many are removed. 

In the 'xyls' file there are also many locations which in the picture are blocked by the ground or buildings, so those stars could not possibly have been detected from the picture. So how can the software determine xy positions for those stars?

Also, sometimes some stars which are clearly visible in the picture, are marked in the 'objs.png' picture but are missing from the 'xyls' file. Why is this? 

Can you tell me the relation between the indicated locations in the 'objs.png' image and the locations in the 'xyls' file?

[Dustin Lang]

Hi,

There are two xyls files - one will be named like IMAGE.xyls ("stars" detected in your image), the other IMAGE-index.xyls (known stars from the index, projected into your image space) - which one are you talking about here?

If you can send an example pointing out one or two discrepancies, that should be helpful to sort out what is going on.

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/faf747f6-b684-490c-9894-4687ae81cf60n%40googlegroups.com.

[Dustin Lang]

(answering Max's private follow-up message on-list...)

Yes, IMAGE-indx.xyls will be the "index" stars projected into your image.

I mis-spoke before, the X,Y positions of "stars" we detected in your image are in the file called IMAGE.axy (axy = "augmented xy list" -- it just has a few headers describing the astrometry.net job to run).

Oddly, solve-field doesn't automatically produce the list of stars detected in your image projected into RA,Dec!

But you can convert between an X,Y list and an RA,Dec list by using the wcs-xy2rd and wcs-rd2xy programs --

wcs-xy2rd -w IMAGE.wcs -i IMAGE.axy -o IMAGE.radec

There is also the corr.fits file or IMAGE.corr, which lists stars that we think matched - with their image X,Y, reference RA,Dec, and the conversions each way.

cheers,

dustin

[Max van de Kamp]

Hi Dustin, Thank you for your message!

I get only the file IMAGE-indx.xyls (without 'e' in 'indx'). But your answer seems to explain the discrepancy, if that is the file with stars from the list projected onto my image.

How can I get the file IMAGE.xyls too?

And is there also a file with the right ascension and declination of the stars detected in the image? Or are those stars not identified yet at that point?

Also, I didn't know the code makes a projection of the star space onto my image space. Is there a way to output also the entire xy-matrix of the image according to that projection? That would be useful.

Best regards

Max