iraf.net

I am doing photometry on 90-300 stars per frame. My intent was to use the ccmap/xy2sky packages to create a coordinate list of these stars and then use those coordinates to match my stars to themselves in the USNO database. After painstakingly creating an input file for the ccmap package and running this package, the fitted coordinates returned by ccmap are sometimes dozens of arcseconds off of the, as far as I can tell, accurate coordinates that were input. e.g.:
19:34:32.367 45:53:39.75 19:34:21.117 45:53:05.22 117.44 34.535
Here, the first two columns are the RA/DEC I used as input in the ccmap file, the middle two columns are the coordinates ccmap fitted to the input, and the last two columns are the residual RA/DEC, which you can see is quite high.
This causes major fitting problems when trying to match my stars to the USNO database. I have tried using ccmap interactively and changing the fit order, etc. did not seem to have any effect on the residual values. Is there any solution to obtain a more accurate fitting?

I use ccmap and the USNO catalog all the time with accuracy matching that of the catalog source positions. If you would like to send me (valdes@noao.edu) a coordinate file, a parameter listing (lpar ccmap), any output you see, and anything else needed to do what you are doing I can probably pinpoint the problem.Frank Valdes