One property of FIGARO and its derivative OBSERVING is that files are
never deleted automatically. Thus, if you flat field 100 frames, you
will end up with the original 100 frames plus 100 flat fielded frames.
This means that by the time you leave, or perhaps even at the end of a
night, you may be close to running out of disk space. To check the
amount of available disk space, type FREE. When you start your run,
you can use the command DCLEAR, which will delete all non-system
files. DCLEAR is a very dangerous word, as it will delete everything
that has been written to the scratch disk. Do not use it unless you
mean to. (If you type it by mistake, don't worry - it asks the user
for confirmation - twice.)
The new CCD controller differs from the old one in several respects.
In addition to supporting four cameras, there are eight binning
patterns per dewar which can be remotely controlled. (Note that only
the newer CCD dewars contain the necessary hardware as part of their
saddlebag electronics to enable remotely operated binning to function.
The software is aware of which are so configured.) Also, there is a
continuous temperature display which is updated every three
minutes. (Temperature display begins only after the command CAMERAS
has been executed.)
A note on nomenclature -- the CCD controller has four ports, and thus can run four CCD chips simultaneously. We refer to these four ports as the four cameras. These are numbered 0-3. Plugged into any of these four cameras can be one of many dewars. The numbering system for the dewars is described in the memo by Fred Harris which is Appendix A. Thus, when an astronomer begins a run (or changes instruments), one must first inform the computer how many of the four controller ports are actually in use, and which port numbers are in use. Next, the computer needs to know the dewar number that is plugged into each port in use (i.e., what type of CCD is in each camera and what is its default readout pattern). We have stored a file with default values for each of the standard CCD dewars (the numbering scheme is the usual one, so that dewar 3 refers to the dewar containing RCA CCD which is normally used used for direct imaging at the 60-inch telescope). The word which conveys all this setup information is CAMERAS. CAMERAS asks a series of questions which you should answer. The first question is what telescope you are at (60 or 200 inch).
CAMERAS then presents you with the default values for each of the dewars you have indicated are in use. These default values are of two categories: some cannot be modified while others can be modified by any observer. Those which cannot be modified include the chip size in pixels, and other parameters which are fixed once the CCD chip is defined as RCA, TI, or Tektronix. The parameters which can be modified are which part of the image is to be saved and the binning pattern that is desired (for those dewars where remote control of binning is supported). The part of the image to be saved is normally the whole image in direct imaging, but may be only part of the image for spectroscopic frames. There is also, for enginering purposes only, a choice of dewar 0, where all parameters can be modified.
The dewar information will indicate whether or not remotely controlled binning is supported on that dewar. If it is, consult the list of binning patterns for that dewar and choose the one you desire. The binning patterns are listed in Appendix A. The specification of the area to be read is given by the starting position in x and y and by the number of pixels to be saved in x and y. The former is defined in terms of real pixels on the CCD, independent of the choice of binning pattern. The number of pixels to be saved is given as actual pixels readout (binned if binning has been chosen).
Spectroscopic observers should next type the word ANGLES. This asks a
series of questions about what grating is in each camera and the
grating angle. These questions must be answered for each camera in
use. If you change gratings or their angles during the night, you
must go through ANGLES again. Otherwise, the subsequent headers will
not have the correct grating angles.
Direct observers should next type SCALE. This word asks you for the
image scale (arc-sec/pixel). (If you are using on-chip binning, the
scale should be arc-sec/binned pixel.) A number of suggested values
for various situations are typed out. The image scale is only used in
calculations of seeing, so it is not critical that it be absolutely
correct. Suggested values for the P60 are for the RCA CCD, no
reimaging -- 0.47 arc-sec/pixel, RCA ccd with reimaging -- 1.2
arc-sec/pixel.
Most observers have their own system of night numbers. The night
number can be set by typing the command NIGHT.
This word allows you to set the frame counter, which determines the
name of the image, to an arbitrary value. There is a separate counter
for each of the exposure types listed in appendix C. The protocol for
image names involves the current value of the frame number and is
described in Section V.
The first thing done by CHIPORIENT is that a two segment short
exposure is taken. Between the two exposures, the telescope is moved
x arc-sec N, where x is a number corresponding to 1/3 of the
number of rows in the frame multiplied by the image scale (from
SCALE). Next, the frame is displayed and you are asked to identify
pairs of images of the same star with the cursor; after a pair is
identified, the angle between the NS line and the columns is
calculated and typed on the terminal, as is the scale in
arc-sec/pixel. This number should then be stored as the image scale
(use the word SCALE). You can repeat this for several pairs of stars
if desired without redisplaying the frame.
After taking an arc exposure, display it on the monitor using either
IMAGE or SIMAGE (see section XI). SPECORIENT asks you to put the
cursor on one side of the slit on an arc line, then on the other side
of the slit. Over a series of 10 pixel wide intervals along the
length of the slit, SPECORIENT calculates (via a Gaussian fit) the
centroid of the arc line, the summed maximum intensity with the
background subtracted off, and the full width at half maximum of the
spectral line. The intent is to use SPECORIENT to align the slit with
the rows of the detector and to check for focal variations along the
slit. SPECORIENT will fail if the slit is grossly tilted (more than
15 pixels across its length).
OBSERVING assumes that in raw spectroscopic frames, the dispersion
direction is along the columns (i.e., vertical). This is in fact
always the case in all Palomar instruments that use CCD detectors in
order to minimize the effects of charge transfer on spectral
resolution. (Previous software at the 200-inch rotated the spectra so
that the dispersion was along the x direction of a frame before the
astronomer ever saw the data.) FIGARO assumes that the dispersion is
along the x-direction of a frame, i.e., the opposite of OBSERVING.
The necessary 90
rotation can be achieved using the FIGARO
command ROTATE, but should only be done if you plan to use one of the
more exotic spectral analysis features of FIGARO, for example
correction of s-distortion or fitting a 2 dimensional wavelength
solution.