Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.sao.ru/precise/Midas_doc/doc/94NOV/vol2/node159.html
Дата изменения: Fri Feb 23 14:01:48 1996
Дата индексирования: Tue Oct 2 18:29:21 2012
Кодировка:
Поисковые слова: п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п р п р п р п р п р п
|
Next: Retrieving demonstration and
Up: Echelle Reduction Method
Previous: Echelle Reduction Method
The information involved in a reduction session consists of user data
and system tables. User data is a set of echelle images, observed with
the same instrument configuration, including a wavelength calibration
image ( WLC), a flat field image ( FLAT) and astronomical
objects OBJ. Optionally, this set will include standard stars
( STD) to be used in the absolute or relative flux calibration,
and dark images ( DARK). Catalogues with comparison lines and
absolute fluxes for standard stars are available in the system as
MIDAS tables.
Before starting the actual reduction some preprocessing of the data is
required to correct for standard detector effects as follows:
- Rotation of input frames.
After this rotation,
the dispersion direction of the echelle orders will be horizontal,
with wavelengths increasing from left to right and spectral
order numbers decreasing from bottom to top of the image.
As always in MIDAS, the origin is the pixel
,
located in the lower left corner of the image.
- Updating START and STEP descriptors.
Descriptors START and STEP must be set to 1.,1. for
all images processed. Session keyword CCDBIN must be set
to the original binning factor along x- and y-axis. Image
rotation and descriptors update are performed by the command
ROTATE/ECHELLE.
- Cleaning of bad columns.
First, bad columns -- bad rows after the
rotation -- can be removed with the command COMPUTE/ROW.
The cleaning of bad columns is required for FLAT images where
the variation of the intensity due to these columns can affect
the automatic detection of the orders.
- Cleaning of hot pixels.
Hot pixels can be eliminated by filtering the images. In
case the observation has been splitted in several exposures
and more than one image is available with the same information,
the images can be averaged with the command AVERAGE/WINDOW; this
command can, optionally, interpolate pixel values with large deviations
from the average value. Removal of hot pixels is required for
DARK images and is recommended for OBJ exposures.
General methods to clean bidimensional spectra are described in
Chapter
(Removing Cosmic Ray Hits). The
command FILTER/ECHELLE, adapted to echelle spectra is described
in Section
- Subtraction of dark current from FLAT, OBJ and STD
frames.
The dark level is estimated from a series of DARK exposures
of short duration which are averaged to reduce the effect of the
read--out
noise of the CCD and to eliminate hot pixels as described before.
If preflashing is necessary, a set of preflashed
DARK exposures should be obtained in a similar manner. It is
advisable to obtain a set of DARK images with similar exposure
times as the object and standard star frames, or to scale the dark
level to the observed exposure.
- Checking exposure times in OBJ and STD frames. For images
generated by ESO instruments, the exposure time (in seconds) is stored in the
descriptor O_TIME(7). If necessary this descriptor can be
created as O_TIME/D/1/7 with the command WRITE/DESCRIPTOR.
Next: Retrieving demonstration and
Up: Echelle Reduction Method
Previous: Echelle Reduction Method
Pascal Ballester
Tue Mar 28 16:52:29 MET DST 1995