Matches only in a section of the image

[Max van de Kamp]

Hi Dustin,

I'm using astrometry for pictures which have a wide angle view, often 60 or 90 degrees wide or even more. 

I usually use the  option '--scale-low 45' on the command line. At least I found it works a lot better that way than if I don't specify the scale.

Often the result is good, but sometimes the result is a bit strange. When I check from the file '.corr' which stars were matched, sometimes those stars are all grouped in a section of the image. And if for instance all those matched stars are in the left-hand side of the picture, the mapping goes a bit wrong on the right-hand side.

What can be the cause of this? Should I maybe specify the scale more exactly at what I think the angular size of the picture is? Or should I crop the picture to partial images?

Best regards

Max

[Dustin Lang]

Hi,

This happens because, like you saw, we matched a quadrangle of stars in one part of your image to the reference catalog, but then when we tried to fit the distortion, some of the stars were too distorted for it to latch on to them.

Setting the scale won't help - that just affects how fast the solution is found.

You can try setting the "CRPIX in Center" option -- this forces the center of the project to be the center of your image, which usually helps.

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/25c5d338-c843-4a55-987e-99480519be98n%40googlegroups.com.