Status: offline
Registered: 09/30/2005
Posts: 4040
|
I have recently got IRAF v2.13 installed on my computer to deal with fits files that are written in ZPN projection and also to manage images from the Spitzer Space Telescope. I notice that IRAF does an erroneous job without complaining with both of these images.1) When I try to use the task "imcombine" to combine multiple images with ZPN projections and ask it to use wcs for combining, it produces an image that is not correctly combined (infact totally messed up) but the task runs without complaining. Prior to installing this new version of IRAF, the task did complain that it can not handle the ZPN projection.2)Spitzer Telescope images for the GLIMPSE program are written in GLON-CAR, GLAT-CAR projection which can be viewed correctly by ds9. However, if I use the imcctran task to convert the images to FK5 format, the task runs without complaining but does a bad job. It forcibly rotates the wcs axis to pixel xy axis and messes up the coordinates.Are these bugs ?!
|
Status: offline
Registered: 09/30/2005
Posts: 4040
|
There may well be bugs, but these at least aren't known issues at the moment. The ZPN projection code came from a patch being distributed by the Cambridge Astronomical Survey unit and I'll admit wasn't rigorously tested. The CAR code has been in the system for a while and aside from an old known bug involving a degree/radian conversion, is also assumed to be working properly. I checked the ZPN code for the same error but didn't see anything obviously wrong.As for the CAR code, use the banner box to search the forums for old mail on "GLON-CAR", there are only a few messages but they have a useful discussion of the problems of imcctran and this class of projection. The imcctran in V2.13 should use the longpole/latpole method discussed but I'd have to see an image or have more information to comment further.In the case of ZPN and IMCOMBINE, do all the images have a ZPN projection or are there mixed projections? Can you supply an example data file and the parameters you're using (even a long header listing would help).Cheers,
-Mike
|