MPE-3D Pages
|
MPE-3D Special Notes for Specific Runs:October 1998On this run the halogen lamps which we used for the dome flatfields had very different spectral characteristics than those used in previous (AAT) runs. They were not as spectrally smooth as former lamps, rendering them unusable for the spectral component of the (at least) K-band flatfield. Fear not! This simply means that to generate proper flatfields for this run you must do it the "old-fashioned" way.What this means is that you will use the dome exposures for the spatial
component of the flatfield and the
incandescent
bulb exposures for the spectral component. The only difficulty in doing
this lies on my side. That is, since in this case each of the individual
256 x 256 must be individually wavelength calibrated, I had to generate
the appropriate calibration values for these images as well as the merged
ones. I have done so for the case of the K-band values, but have not yet
determined the necessity of doing it for the H-band (I think it is not
an issue). So, without further ado, here is a sample COLA script
(red_flat_os.col)
for generating appropriate flatfields:
The calsets in the above script were made with this script
(red_cal_os.col):
November/December 1998As the dark exposures for this run were taken on the final morning, they were not available at the time when a number of observers wrote their data to tape. Since the run, I have "post-pre-processed" all the dark files (that is, I've already run the ".gs" files through 3dcmean to produce cleaned average frames). These files are available in the Darks-NovDec-1998 directory. For each dark exposure there is a gzip'd tar file of gzip'd ".descr" and ".image" files. For example, the file dark_15s.tar.gz contains the files for the 15 second dark exposures. To extract them from that file follow these steps:
March 1999
So, a few changes had to be made to the software to accomodate data taken in this fashion. The modified code includes 3dlin.col, 3dcenter2000.col, 3dstar2.col, 3dcubin.c, 3drebin1.c. NOTE: Also be sure to check the Data Reduction page for non-piezo-related updates to other routines. In the case of 3dlin.col, the change involved introducing the concept of using a value of 0 for the Keyword PIEZ. This value is recognized when using values of 3, 33, 5, and 55 for the Keyword SEQU. In those cases the result is a series of on-off pairs which are sequentially numbered, without the "customary" p1 or p2 appearing. See the top of the script itself for details. In the cases of 3dcenter2000.col and 3dstar2.col the changes involved introducing an automatic means of determining whether the data were taken in the normal fashion, or taken without the aide of 3D's piezo. In the latter case simply enter the same name for the Keywords SET and SETP. See the tops of the scripts themselves for details. Further, if you have data taken without the aide of 3D's piezo the wavelength calibration is done within these scripts using starcal2 rather than 3dcal (so be sure to get a copy in the usual place). Finally, I took this opportunity to make fundamental changes to 3dstar2.col such that it more resembles the structure and function of 3dcenter2000.col (by adding the Keywords TWEAK and DPCOR). In the cases of 3dcubin.c and 3drebin1.c the changes involved enable the routines to operate normally on either 256 x 600 (normal, piezo subsampled) data or 256 x 300 (no piezo) data. In each case the routines autosense which kind of data is being input and act accordingly. In addition, as always there is a new input file (3dcubin.991k) for use in reducing K-band data from March 1999. This page was last modified on 06April 1999.
|