iraf.net

Hello,

I am using msctpeak for tweaking the WCS on various images---some from mosaic detectors, others from single-chips. msctpeak is really convenient because of its nice interactive features (vs using ccfind and ccmap, for instance).

My problem is that when I try to use a non-linear plate solution (tnx projection and general fit), I often get a reference point that is located far off the actual image. As a result, the astrometry is not so great where I actually need it. Is there some way I can force msctpeak to set the reference point (crpix1 and crpix2, or crval1 and crval2) where I want it---or at least in the same vicinity? I have no clue how it actually picks what point to use, since this is not an input parameter.

Thanks very much!
Frank Winkler