Re: Star.exe help

["John Phillips" <jdp@indy.net>]

Thanks for you insight Mike.  I think the problem may be in your words "star
uses Mark III sized sections of the catalog" to try to match.  I had
concluded that star used a CDELT1*NAXIS1,CDELT2*NAXIS2 sized chunk of the
catalog to try to match.  I will do some more experimentation.

The image I am working with has stars of all magnitudes brighter than a
limit  I choose.  I went down to 15.0.  At about magnitude 7 or so, it
starts making larger, in addition to brighter spots to represent the stars.
It seems pretty realistic in that regard.  Will this break the selection
algorithm star uses?  I was using a larger than Mark III section of the sky,
so there were several brighter stars.  I think they were saturated though.
Is this important?

Thanks for your help Mike.

John Phillips
jdp@indy.net

-----Original Message-----
From: mgutzwiller@lanvision.com <mgutzwiller@lanvision.com>
To: tass@wwa.com <tass@wwa.com>
Date: Tuesday, March 16, 1999 9:32 AM
Subject: RE: Star.exe help


>A couple of things come to mind for the catalog mismatch. The match
>routines try to match the N brightest stars in an image to the N brightest
>stars in the catalog where N is the "Number of Stars to Match" in the
>options dialog. If the image's N brightest unsaturated stars don't
>correlate to the N brightest stars in the catalog sections star uses the
>ctalog match will fail. This could be caused by a several things:
>
>* Different image size - star uses Mark III sized sections of the catalog
to
>try to match. If the image is significantly different in size the match
>will fail.
>* Different saturation levels - the tyc_good catalog has the brighter stars
>deleted as well as stars that would be too close to resolve in a Mark III
>image. If your artificial image has no such limits the match may fail.
>* If the PSF of the generated stars is not consistent, i.e. brighter stars
>are bigger, not just brighter, star will not measure the magnitude
correctly
>and might not have the correct N brightest stars.
>
>Hope this helps,
>
>Mike G.
>
>
>-----Original Message-----
>From: John Phillips [mailto:jdp@indy.net]
>Sent: Monday, March 15, 1999 9:21 PM
>To: tass@wwa.com
>Subject: Star.exe help
>
>
>I can sense the excitement in the group as Tom's gets ever closer to making
>the Mark IV work. I am anxious to see the data from these first shots.
>
>Please, have mercy on a beginner, as he changes the subject back to
>something you all learned long ago!
>
>I have been working with Star.exe and have been able to successfully
process
>all of the images from the 'archive of early images' page. I then used my
>skymap program to create a fts file of stars near the equator. When I ran
>Star.exe on this file, it didn't find any stars at all. It seems that the
>'sky' didn't have any noise in it, so the routines using sigma didn't work.
>I wrote a little program to add gausian noise to each pixel. Now Star.exe
>can find all of the stars in the image, but can only find the x,y
>coordinates of the stars. I get the dreaded 'match to catalog failed'
>comment in the generated file. I think I have the CTYPEn, CRPIXn, CDELTn,
>and CROTAn keywords set up properly to point Star.exe to the right portion
>of the sky. At least the generated RA and DEC are very nearly right for
>these stars. Still, I would get a warm fuzzy feeling if I could make the
>match work.
>
>Does anyone have any clues to what may be going wrong? Any comments are
>welcome.
>
>Thanks,
>
>John Phillips
>jdp@indy.net
>
>

smime.p7s