iraf.net

After successfully reducing a large number of data sets today using CCDPROC it has begun to act strangely. Every time I try to process a group of files and apply a flat field which I have used imsurfit to remove a trend, the flat field image is being overwritten after processing the first image. The resulting flat field is either all 1's or just random noise centered on 1. (The starting flat image was scaled to a mean of 1 when it was created by using imsurfit with the "response" option). As I mentioned, this was working fine all day. The only thing I did near the end of the day was to use imsurfit to fit a second order surface to my flat and create a response file which I was then trying to use as a flat. I had already done this earlier with a first order fit and it worked.The problem appears to be related to the fact that the CCDMEAN value in the header file is 16766 but the image data is normalized to 1. Therefore when ccdproc goes to process the data it scales the data by 1/16766 and then it appears to try to save over the flat file. Here are my questions:1. I understand that in order to process the FLAT data it is necessary to divide by the mean but why is it corrupting my existing flat file in this one case? When I use a regular flat file which has not been fit this does not happen. Nor does it happen when I use the file with the linear trend removed.2. Somehow I managed to use imsurfit to remove a linear trend earlier today and the CCDMEAN value was saved correctly as "0.9999989". I cannot seem to get that to happen again. Do you know what I am doing wrong? By the way I do not seem to be able to recreate my linear fit file with the correct CCDMEAN value. I have no idea how I did this.I am sorry if this is confusing and a little long but this has me puzzled.

CCDPROC uses the CCDMEAN and CCDMEANT keywords to scale a flat prior to applying it. CCDMEANT is like a time stamp which is compared to the modify time in case some non-ccdred task was used as in your case. Deleting these keywords can be used to affect things. If CCDMEANT is removed then the CCDMEAN value will be used regardless of the modify time of the image. If CCDMEAN is removed then CCDPROC will recompute the value and store it in the header (which modifies the image in the sense of the header). Some, though not many, non-ccdred tasks will remove CCDMEANT and/or update CCDMEAN for you. But you should probably check this.If what you are doing normalizes the data then you should delete CCDMEANT and set CCDMEAN to 1 using HEDIT.As far as modifying the file, CCDPROC will process calibrations if it determines that some processing step was not already performed. I'm not sure what would have happened with your data. Any processing should be reported in the logfile or output to the terminal.I hope this addresses your problem.Yours,
Frank Valdes