Welcome to iraf.net Wednesday, April 24 2024 @ 05:08 AM GMT
Buglog #474
- Thursday, December 07 2000 @ 09:22 PM GMT
- Contributed by: admin
- Views: 1,199
MODULE: mscred.ccdproc
SYSTEM: -V4.1: December 5, 2000
SYSTEM: -V4.1: December 5, 2000
NUMBER: 474 MODULE: mscred.ccdproc SYSTEM: -V4.1: December 5, 2000 DATE: Thu Dec 7 14:22:10 MST 2000 FROM: valdes BUG: When a flat field is processed in a certain way as noted below, applying it to object data with CCDPROC will cause each flat field extension to be normalized differently. This will be evident in the log file where the scale factor for the flat field extensions will be listed with different values rather than the same value. The flat fielded object image will then appear with different gain levels of the same relative magnitude as the flat field in each extension. Note that small residual gain differences are normal due to dome flats not exactly matching the sky and these are removed by using sky flats. The incorrect behavior occurs when the flat field is processed after combining to a master flat field followed by making a new version of the flat field, such as when making a pupil ghost corrected flat field. The error occurs because a time stamp for the CCDMEAN scaling keyword, the keyword CCDMEANT, is added by CCDPROC and checked against the time the image was last modified. After the first processing all the CCDMEAN keywords will correctly have the same value. However, any modified version of the flat field will then invalidate the time stamp causing the next execution of CCDPROC that applies the flat field to compute new CCDMEAN values separately for each extension. If the CCDMEANT keyword is absent then CCDPROC will not recompute the normalization value. Thus the workaround is to delete the CCDMEANT keyword from all extensions either before making a new version of the flat field or in the flat field version to be applied by CCDPROC. In the most common case where individual flat fields are processed with scaling followed by combining to a master flat using the task FLATCOMBINE the bug does not occur because the scaling and combining operation eliminates the CCDMEANT keyword. Also when the flat field is not modified after it has been processed by CCDPROC the error will not occur. So almost all users will not have encountered this problem. STATUS: The MSCRED release of December 7, 2000 fixes this problem by having CCDPROC eliminate the CCDMEANT keyword when the flat field is first processed.
The following comments are owned by whomever posted them. This site is not responsible for what they say.